Using actively fine-tuned machine-learning interatomic potentials with experimental validation, the authors report that cation disorder improves lithium-path connectivity but broadens site energies, suppressing diffusion in disordered-rocksalt cathodes. They develop an energy-adjusted percolation descriptor and a Li–Mn–Ti–O design map. This is a composition-specific transport result; the review does not assign individual capacity figures to modeled or measured material states.
Technology: Actively fine-tuned machine-learning interatomic potentials
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Publisher abstract describes atomistic analysis combined with experimental validation within the discovery study.
- Still unknown: Review scope was the publisher abstract and metadata; full experimental details and capacity attribution were not assessed. Article-in-press text may change.
- AI role: Learned atomistic potentials enabled analysis of lithium connectivity and site-energy distributions.
Evidence sources: https://www.nature.com/articles/s41467-026-79420-w
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
An AI-assisted 3CL-protease inhibitor study selected 249 molecules from about five million generated candidates and evaluated 18 using free-energy perturbation. Five molecules were experimentally tested; all five inhibited the target below 1 micromolar IC50, and one inhibited more strongly than its lead. MMELON activity predictions had Pearson R=0.72 in a blind test. This summary is limited to the abstract and November 13, 2026 publication metadata; it does not establish antiviral efficacy in people.
Technology: MMELON biomedical foundation model with generative molecular models
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Five synthesized molecules tested for 3CL-protease inhibition after computational candidate generation and filtering.
- Still unknown: Source checking does not establish independent confirmation. Review date: 2026-11-15.
- AI role: Generative models produced candidate molecules; the MMELON foundation model predicted activity and supported downselection.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=EXT_ID%3A43024107&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The November 13, 2026 paper introduces WenHai-E2E, which generates global eddy-resolving ocean forecasts directly from observations and reports better forecast accuracy than a numerical system. The abstract and publication metadata reviewed here identify no separate oceanographic discovery or specific field result beyond forecast performance. Full text was not assessed.
Technology: WenHai-E2E (end-to-end deep learning ocean forecast model)
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Forecast accuracy against a numerical ocean forecasting system, without a separate discovery in the reviewed abstract.
- Still unknown: Admission review pending: primary full results needed to distinguish a specific scientific/application result from general method validation. Existing supported claims may be retained with bounded scope, but do not mark source checked until admission is resolved.
- AI role: Machine learning model produces the full eddy-resolving global ocean forecast from observations, replacing numerical initial conditions.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.1126%2Fsciadv.aeh9058%22&resultType=core&format=json
Needs source check · editorial details
The source has not been fully checked or contains conflicting information. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
MIT’s November 13, 2026 announcement describes a machine-learning copilot that flags regions for human edits during topology optimization. It reports improved manufacturability and buckling-load performance, but the reviewed announcement identifies no separate structural discovery or experimentally demonstrated new structure beyond testing the design workflow. Synthetic feedback training and limited thickening/thinning edits are acknowledged. The full journal paper was not assessed. The publisher abstract specifies improved manufacturability or a 39% linear-buckling-load improvement; it does not assign 39% to both metrics. The paper was published November 10; this entry represents the November 13 announcement.
Technology: Human-in-the-loop AI copilot with topology optimization (machine-learning region-flagging model)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Computational topology-optimization workflow demonstrations; no separate structural finding established in reviewed scope.
- Still unknown: Admission review pending: primary full results needed to distinguish a specific scientific/application result from general method validation. Existing supported claims may be retained with bounded scope, but do not mark source checked until admission is resolved.
- AI role: A machine-learning model quickly identifies design regions a human would likely want to change, within a topology optimization workflow.
Evidence sources: https://news.mit.edu/2026/new-ai-copilot-helps-engineers-design-better-structures-1113 · https://www.nature.com/articles/s44172-026-00818-x
Needs source check · editorial details
The source has not been fully checked or contains conflicting information. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
In the journal abstract, an AI-ECG model trained on more than 330,000 ECG–echocardiogram pairs detected the dilated-cardiomyopathy phenotype (AUROC 0.87–0.97) and predicted later disease (C-index 0.80–0.87). In familial cohorts, five-year negative predictive values were 95.5% in Antwerp and 91.5% in Maastricht; positive predictive values were 42.4% and 39.4%. These are diagnostic and risk-prediction findings, not demonstrated clinical benefit.
Technology: Explainable deep learning model applied to electrocardiography
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Validated in UK Biobank, Zhongshan Hospital and Antwerp cohorts, and in two familial cohorts (Antwerp n=166, Maastricht n=177), compared with traditional ECG features and echocardiographic traits.
- Still unknown: Review limited to the indexed abstract and publication metadata; full-paper methods and additional limitations were not assessed.
- AI role: Deep learning model trained to detect and predict the echocardiographic DCM phenotype from ECG; explainable AI-ECG outperformed traditional ECG parameters.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.1093%2Feurheartj%2Fehag917%22&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The Germinal Multi preprint abstract reports antibodies designed against several antigen variants at once. Experimental screens of tens of designs per campaign found binders to resistance-associated EGFR variants and neutralizers of historical SARS-CoV-2 variants, including variants withheld from design. Redesign of BD55-1205 restored neutralization of a forecast-guided engineered escape variant while retaining parental-variant activity. The escape result concerns an engineered variant, not a demonstrated future circulating strain. Review is limited to the abstract and original-server metadata.
Technology: Germinal Multi (computational antibody design framework); evolutionary forecasting model
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: Tens of designs per campaign were screened experimentally; designs bound EGFR variants and neutralized SARS-CoV-2 variants, including variants withheld from the design process.
- Still unknown: Review limited to the indexed abstract and publication metadata; full-paper methods and additional limitations were not assessed.
- AI role: Germinal Multi computationally optimized one antibody paratope against multiple antigen variants, and an evolutionary forecasting model guided construction of an escape variant and antibody redesign.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.64898%2F2026.11.10.765734%22&resultType=core&format=json · https://api.biorxiv.org/details/biorxiv/10.64898/2026.11.10.765734
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
MIT researchers with Imperial College London designed a small synthetic set of deliberately unrealistic emissions scenarios to train climate emulators, which normally need six to eight full Earth system model runs that each take months. The method needs only one model run, and the emulator predicted unseen scenarios and local pollutant-control effects where traditional emulators failed. It was validated with a simple climate model rather than a full Earth system model, and the article reports no quantitative accuracy metrics.
Technology: Climate emulator trained on optimized synthetic scenarios
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: An emulator trained on a small set of synthetic scenarios predicted unseen scenarios and a local pollutant-control effect where traditional emulators failed. The source says it was validated with a simple climate model.
- Still unknown: The source says the method still needs testing on a full Earth system model scenario and on other climate model types; it does not say whether this has been done.
- AI role: An optimization algorithm generated the synthetic training scenarios; an AI emulator was trained on them to predict climate outcomes.
Evidence sources: https://news.mit.edu/2026/new-research-could-lead-faster-trustworthy-ai-models-climate-prediction-1112
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
GPT-6 Astra produced explicit mutually orthogonal Latin-square constructions that improve known lower bounds for N(n) at 33 values of n from 22 through 494. The standout construction gives four MOLS of order 22, completing the last open case needed to show N(n)≥4 for every n>10. The same paper reports resulting bounds at other orders and derives a strongly regular graph with parameters (484,126,40,30) from the order-22 construction; these connected consequences are counted once as one paper-level result.
Technology: OpenAI GPT-6 Astra, used through ChatGPT Pro and the OpenAI Codex harness
Family GPT-6: profile · models page
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: GPT-6 Astra produced explicit mutually orthogonal Latin-square constructions that improve known lower bounds for N(n) at 33 values of n from 22 through 494.
- Still unknown: This is an arXiv preprint. The constructions are finite combinatorial objects, checked by code written by the author and separately reconstructed/checked by Claude Opus 5.5 as reported in the paper. Counted once despite 33 parameter values and multiple corollaries.
- AI role: The author states GPT-6 Astra found all listed MOLS constructions and wrote their descriptions/code. The author wrote independent verification code for each construction; the paper also reports an independent Claude Opus 5.5 recreation and verification from the descriptions without the supplied code.
Evidence sources: https://arxiv.org/html/2611.16299
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
SALT spectroscopy of the previously unclassified source LMC tSNE b8 (R.A. 06:22:41.86, Dec. −68:59:21.9) identified it as an active galactic nucleus at redshift z = 2.80. The source was selected from the blazar-like sequence in a t-SNE map combining VISTA near-infrared and ASKAP radio survey data with multi-wavelength photometry. This is one specific spectroscopic classification from the paper’s broader follow-up sample.
Technology: t-distributed stochastic neighbour embedding (t-SNE) applied to multi-wavelength source photometry
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: SALT spectroscopy of the previously unclassified source LMC tSNE b8 (R.A.
- Still unknown: This is an arXiv preprint. The selected source was classified as an AGN at z=2.80 from SALT spectroscopy, but the paper’s broader 18-target blazar-sequence follow-up also contains Magellanic objects and two low-S/N unconfirmed sources; those are not counted as confirmed discoveries. The claim here is limited to LMC tSNE b8. The paper calls the selection sequence blazar-like; that label does not establish this specific source as a blazar.
- AI role: t-SNE mapped 6,652 Large Magellanic Cloud and 1,356 Small Magellanic Cloud sources using optical-to-radio features. The unsupervised map exposed a blazar-like sequence from which LMC tSNE b8 was selected for SALT follow-up; spectroscopy then established its AGN classification and redshift.
Evidence sources: https://arxiv.org/abs/2611.15880
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
GenDIReCT used diffusion-based reconstruction conditioned on calibration-independent closure invariants to analyze EHT observations from 2017, 2018 and 2021. The authors recover ring morphology and estimate a diameter of 42.2 ± 1.5 microarcseconds; geometric modeling gives 42.3 ± 1.1. This is a new analysis of existing observations, overlapping the target and some data of the cataloged 2023 PRIMO reconstruction. It is not a new observation or the first black-hole-shadow discovery.
Technology: GenDIReCT latent diffusion
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Archival observational reconstruction, synthetic validation and geometric modeling; lower-quality epochs give less distinct morphology.
- Still unknown: Sparse interferometric coverage and reconstruction assumptions remain limitations; physical mass estimates require separate scaling assumptions.
- AI role: Latent diffusion reconstructed images from closure invariants.
Evidence sources: https://www.nature.com/articles/s41586-026-11241-9 · https://arxiv.org/abs/2608.19675
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Google Research reported a wearable-based health-age model associated with subsequent hypertension, cardiovascular disease and type 2 diabetes in the All of Us cohort (22,748 participants). A companion world model simulated 12-week trajectories. In 490 blinded paired reviews, coaches preferred simulation-grounded plans in 71.4% of evaluations; physicians rated 97.6% of physiological claims accurate. This summary is limited to the November 11, 2026 institutional blog. The authors state all evaluations were offline and retrospective; clinical effectiveness and durable habit change require prospective trials.
Technology: Coachable health-age model, SensorWM action-conditioned world model, feasibility-constrained optimizer with a language model agent, multimodal CKM deviation model
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: Retrospective clinical-risk modeling and blinded expert comparison of simulated coaching plans.
- Still unknown: Source checking does not establish independent confirmation. Review date: 2026-11-15.
- AI role: Built the health-age metric and multimodal CKM model from wearable and clinical markers, forecast physiological responses to habit changes with a world model, and filtered and converted simulated habit plans into coaching text with a language model agent.
Evidence sources: https://research.google/blog/using-wearables-to-enable-cardiovascular-kidney-metabolic-ckm-risk-profiling-for-personalized-coaching/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Classifiers and UMAP analyzed 129,676 prenatal-test cfDNA samples, with genotype checks in 169 selected samples. The authors report fragmentomic profiles resembling DNASE1L3 R206C homozygotes in some non-homozygotes, while some confirmed homozygotes initially lacked the profile and acquired it in later pregnancies. This extends an already-known genotype association with phenotype discordance and within-person variation; it does not establish the mechanism or clinical utility.
Technology: Supervised classifiers; UMAP
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Existing-cohort analysis with selected ddPCR genotype validation and a small longitudinal subset.
- Still unknown: The validation cohort was enriched for discordance; general-population accuracy and causal interpretation remain unresolved.
- AI role: Classifiers and low-dimensional embeddings identified fragmentomic subgroups and discordant cases.
Evidence sources: https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2026.1887185/full
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The preprint abstract reports that AllositeUltra recovered allosteric sites among its top three predictions in 65.2% of apo structures and 70.6% of substrate-bound structures. More importantly, application to MAGI2 identified a previously unrecognized interdomain pocket, tested by small-molecule screening and X-ray crystallography. This within-study experimental finding qualifies separately from the benchmark. Review is limited to the abstract and original-server metadata.
Technology: AllositeUltra (structure-aware protein language model combined with physical dynamics; residue-guided candidate supplementation)
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: The predicted MAGI2 interdomain pocket was validated experimentally through small-molecule screening and X-ray crystallography; benchmark recovery was measured on apo and substrate-bound structures.
- Still unknown: Review limited to the indexed abstract and publication metadata; full-paper methods and additional limitations were not assessed.
- AI role: A deep-learning model predicted allosteric sites on apo and substrate-bound structures and was applied prospectively to MAGI2 and proteome-wide.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.64898%2F2026.11.09.765238%22&resultType=core&format=json · https://api.biorxiv.org/details/biorxiv/10.64898/2026.11.09.765238
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
TriSCOPE combined cell-level predictive modeling, donor-aware differential expression and pseudotime analysis across six single-nucleus RNA-sequencing datasets (about 4.2 million nuclei from 627 donors). It identified cell-type-specific predictively expressed genes that were often missed by univariate differential-expression analysis; conditional GWAS–eQTL colocalization supported 27 gene–cell-type pairs at Alzheimer’s risk loci, including pairs without significant differential expression. Selected gene signals and analytical branches were compared across independent cohorts.
Technology: TriSCOPE, using LightGBM-based multivariate predictive modeling integrated with differential expression and pseudotime analysis
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Selected gene signals and analytical branches were compared across independent cohorts.
- Still unknown: This is a computational reanalysis of previously published, post-mortem cross-sectional datasets; it did not collect new tissue or perform perturbation experiments. The authors state the framework is for feature discovery rather than clinical prognosis, cannot establish temporal ordering, and requires functional perturbation studies to distinguish pathogenic drivers from protective or reactive signals. The reported cell-level AUCs are modest and should not be presented as patient-level diagnostic performance.
- AI role: The supervised machine-learning branch prioritized gene-expression features that predicted clinical Alzheimer’s diagnosis within cell types, conditional on demographics and APOE genotype. The study integrated those model-ranked features with donor-aware differential expression, pseudotime and GWAS–eQTL colocalization to nominate genetically supported cell-type programs.
Evidence sources: https://www.nature.com/articles/s41467-026-77651-5.pdf
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
In the journal abstract, PREVIA used trauma narratives from 145 patients at two New York City trauma centers to predict PTSD assessed one month later. Reported AUROC was 78.49% ± 5.75%, with precision 46.76% ± 8.66%. On 30 randomly selected patients, its performance was comparable to one PTSD-specialist emergency clinician. The authors describe these as initial findings requiring larger independent-cohort validation; prospective collection does not establish prospective evaluation of a locked model.
Technology: PREVIA clinically informed large language model ensemble
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: One-month PTSD prediction in 145 trauma patients; comparison with a specialist emergency clinician on 30 selected patients.
- Still unknown: Abstract and metadata review only. Authors explicitly call for larger independent-cohort validation; full methods and timing of model selection relative to outcomes were not assessed.
- AI role: LLM ensemble extracts early post-traumatic stress symptoms, risk and protective factors from narratives and produces the risk forecast.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.1038%2Fs41467-026-78768-3%22&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The preprint abstract reports a variational autoencoder trained on natural AAV sequences to generate capsids. More than 255 designs were synthesized and tested in vitro. Seven transduced human cells at over 50% efficiency; four of those avoided neutralization by every AAV-reactive human serum sample tested, at the reported NT50 threshold below 1:50. This establishes in-vitro capsid properties, not in-vivo delivery or clinical efficacy. Review is limited to the abstract and original-server metadata.
Technology: Variational autoencoder (generative model of AAV capsid sequences)
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: More than 255 AI-generated capsid designs were synthesized and tested in vitro; seven transduced human cells above 50% efficiency and four avoided neutralizing antibodies in all reactive human sera.
- Still unknown: Review limited to the indexed abstract and publication metadata; full-paper methods and additional limitations were not assessed.
- AI role: A variational autoencoder learned AAV sequence constraints from natural sequences and generated functional capsid proteins de novo, with a second design campaign exploring uncertain regions.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.64898%2F2026.11.06.764881%22&resultType=core&format=json · https://api.biorxiv.org/details/biorxiv/10.64898/2026.11.06.764881
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A preprint reports external validation of an ECG-image ensemble for Takotsubo syndrome versus anterior STEMI in 95 participants: accuracy 0.914 and AUC 0.970. Among 70 readers, estimated diagnostic correctness rose from 0.632 unaided to 0.864 with AI. Only the fourth phase randomized assistance conditions; AI-assisted reader accuracy remained below model accuracy. This summary is limited to the abstract and version 3 preprint dated November 8, 2026; the first version was posted October 19.
Technology: Self-supervised deep learning ensemble with Grad-CAM explanations
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: ECG-image external validation and a separate four-phase reader experiment; the fourth phase randomized assistance conditions.
- Still unknown: Review limited to abstract and original-server version/corresponding-institution metadata; full paper and historical v1/v2 contents not assessed. DATE/ID MONTH CONFLICT: publication becomes 2026-11-08 while stable ID retains 2026-10.
- AI role: Ensemble classified 12-lead ECG images; its predictions and Grad-CAM maps were shown to readers.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.64898%2F2026.10.15.26365116%22&resultType=core&format=json · https://api.biorxiv.org/details/medrxiv/10.64898/2026.10.15.26365116
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The accepted-paper abstract reports that docking, CDK-liability filtering, machine learning and error prediction guided more selective CK2 inhibitor leads. The authors report submicromolar hits and a 90% pipeline hit rate; kinase-assay figures support biochemical inhibition. These are discovery-stage compounds, with no demonstrated clinical benefit. The hit rate describes the authors' selected pipeline, not a general drug-discovery success rate.
Technology: Machine learning; prediction-error modeling; molecular docking
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Kinase-assay figures report CK2 inhibition by selected hit compounds.
- Still unknown: Full study methods, independent reproduction and clinical outcomes were not assessed.
- AI role: ML and error prediction prioritized compounds after docking and off-target filtering.
Evidence sources: https://www.nature.com/articles/s42004-026-02268-0 · https://www.nature.com/articles/s42004-026-02268-0_reference.pdf
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Subaru/MOIRCS spectroscopy of four PCFNet-selected candidate regions confirmed four protoclusters at z = 3.316, 3.403, 3.408 and 3.425, across three regions, with galaxy-density peaks more than five times the uniform distribution. Twenty-three spectroscopically identified members are largely on the star-forming main sequence. Model-weighted estimates put star-formation-rate density above cosmic averages, while metallicity differences are marginal in two systems and consistent with the field in the others.
Technology: PCFNet point-cloud deep-learning protocluster finder
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Subaru/MOIRCS spectroscopy of four PCFNet-selected candidate regions confirmed four protoclusters at z = 3.316, 3.403, 3.408 and 3.425, across three regions, with galaxy-density peaks more than five times the uniform distribution.
- Still unknown: This is a small follow-up sample: four of 121 PCFNet candidates were targeted, and the parent catalogue is trained on simulated galaxies. Four protoclusters are confirmed across three of four observed regions; a fourth observed candidate region had no significant redshift peak. One of the confirmed regions had previously been reported as a photometric candidate, so this is a spectroscopic confirmation/characterization result rather than four wholly unprecedented structures. The halo masses are inferred from member stellar masses, not direct lensing measurements. The elevated SFR densities depend on membership-probability weighting and are sensitive to assumed geometry, volume and methodology. Metallicity excesses are only 1.4–2 sigma in two systems; two others are statistically consistent with the field. The authors’ proposed gas-inflow/recycling explanation is interpretive.
- AI role: PCFNet used broad-band optical photometry and galaxy positions to assign protocluster-membership probabilities and identify candidate regions. The model outputs determined which regions received Subaru/MOIRCS follow-up and prioritized slit targets; their probabilities also informed member selection and the authors’ volume/SFR-density estimator. Spectroscopy then tested those predictions, finding a 40% member detection rate among high-probability galaxies and confirming three of four observed regions.
Evidence sources: https://arxiv.org/html/2609.35471
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Prospective study of 196 people at five US sites, compared with 24-hour ambulatory blood pressure measurement: 65.8% sensitivity and 90.0% specificity for screening, and 78.4% and 90.9% in the confirmatory test (48 participants). A feasibility study; it does not replace the cuff, and many cases are missed.
Technology: AI-PPG-ACC-HTN (photoplethysmography and accelerometer)
Publication: Peer-reviewed paper · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: prospective or independent check
- Tested: A prospective study of 196 participants at five US sites compared smartwatch screening with 24-hour ambulatory blood pressure monitoring: 65.8% sensitivity and 90.0% specificity, with 78.4% and 90.9% in a confirmatory subgroup of 48.
- Still unknown: The authors call it early feasibility research; the smartwatch does not replace a cuff, many true cases are missed, the confirmatory subgroup is small, and generalizability is unclear.
- AI role: A 1-D convolutional network pretrained with self-supervised contrastive learning on PPG and accelerometer waveforms produced embeddings; a logistic regression output a hypertension probability.
Evidence sources: https://research.google/blog/detecting-signs-of-hypertension-using-wearable-sensors/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 4/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Texas A&M AgriLife Research screened 630,003 protein pairs from about 1,100 human mitochondrial proteins with AlphaFold-Multimer and identified 2,895 likely interactions, more than half not previously reported, giving 85 uncharacterized proteins their first probable partners. The paper appeared in Nature Communications on 25 August 2026 and the institutional release on 5 November. Lab experiments in yeast and human cells confirmed several of the partnerships; most predictions remain untested, and the disease links are potential targets, not established ones. Predictions are shared through the MitoMatch web tool.
Technology: AlphaFold-Multimer
Family AlphaFold: profile · models page
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: AlphaFold-Multimer screened about 630,000 protein pairs; lab experiments in yeast and human cells confirmed several of the predicted partnerships, while the release does not say how many of the 2,895 were tested.
- Still unknown: Only several predictions were confirmed, most remain unverified, the interactions are described as likely rather than established, and disease links are potential targets.
- AI role: AlphaFold-Multimer predicted pairwise interactions among mitochondrial proteins; the predictions were released as the MitoMatch web resource.
Evidence sources: https://www.eurekalert.org/news-releases/1150930
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A supervised deep-learning denoising autoencoder exposed previously undocumented short-period S-diffracted wave observations associated with several known ultra-low-velocity zones (ULVZs). It also revealed four Hawaiian-ULVZ S-diff post-cursor records at Japan F-net stations from East Pacific Rise earthquakes, providing ray paths with mean azimuth about 78° from earlier North American paths. These new cross-cutting observations add constraints on ULVZ geometry and frequency/azimuth-dependent waveforms, but the paper does not present a new ULVZ detection or a completed structural inversion.
Technology: Supervised deep-learning denoising autoencoder (DeepSeisDAE)
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A supervised deep-learning denoising autoencoder exposed previously undocumented short-period S-diffracted wave observations associated with several known ultra-low-velocity zones (ULVZs).
- Still unknown: This paper reports new seismic observations sampling known ULVZs; it does not establish a new ULVZ structure. The Japanese post-cursors are difficult to separate from the main S-diff phases and require future 3D modelling to determine their interpretation and resolve trade-offs. Denoiser amplitude fidelity is a limitation, especially under training/application mismatch; the authors advise caution for single-station denoising. These are observational constraints, not a completed inversion of shape or composition.
- AI role: The denoiser was trained on synthetic seismic waveforms mixed with real global noise, then used to separate signal from noise in real S-diff record sections. Its output helped researchers identify short-period post-cursor observations not reported in earlier literature, including four events recorded in Japan with nearly perpendicular ray geometry to previous North American observations of the Hawaiian ULVZ.
Evidence sources: https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2026JB034067
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The accepted-paper abstract reports that DeltaCata, a geometric graph neural network predicting mutation effects on enzyme kinetics, guided engineering of methyltransferase sGmOMT2. The authors report up to a sevenfold improvement in catalytic efficiency for in vitro icaritin synthesis. This experimental application goes beyond the model benchmarks; industrial-scale performance is not demonstrated by the reviewed abstract.
Technology: DeltaCata; Siamese geometric graph neural network
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The abstract reports experimental enzyme engineering and an in vitro catalytic-efficiency improvement.
- Still unknown: Full experimental methods, industrial performance and independent replication were not assessed.
- AI role: The model predicts mutation-induced kinetic changes to prioritize enzyme variants.
Evidence sources: https://www.nature.com/articles/s41467-026-79287-x
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Claude searched protein databases, designed DNA constructs, reduced purification from about three days to about 90 minutes, and processed cryo-EM data into an initial model in 34 hours. The result is a particle made of 96 copies of a bacteriophage protein, with about 25% more internal volume than the mammalian version. Its biological function is unknown; this is a preprint, and Claude made errors that humans corrected.
Technology: Claude
Family Claude: profile · models page
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: The 3D structure was determined by cryo-EM; cryo-EM confirmed the vault shape for four of seven particle-forming proteins, and fifteen altered protein versions were tested, with a reviewer checking the structure.
- Still unknown: The biological function is unknown, vaults were made only in E. coli, and an unidentified pocket molecule resembling coenzyme A is unconfirmed.
- AI role: Claude searched protein databases, designed DNA constructs, optimized purification, and ran RELION cryo-EM processing to an initial model in 34 hours; it also made errors that humans corrected.
Evidence sources: https://www.anthropic.com/research/investigating-vault-particles-with-claude
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
An agent using Cheminformatics Skills for Antigravity collected thousands of compounds analogous to crizotinib and lorlatinib, analyzed 568, and proposed new rigidified macrocycles with synthesis routes in about an hour. The proposals are computational hypotheses that still need to be synthesized and validated.
Technology: Google Antigravity, Cheminformatics Skills
Publication: Announcement only · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: retrospective or benchmark
- Tested: The source describes computational evaluation only: benchmarks of several machine learning architectures on four readouts with scaffold splits, and predicted potency; no synthesis or experimental validation is reported.
- Still unknown: Every proposed design must still be synthesized and validated; the data are sparse and assay noise is high; the designs are testable hypotheses, not finished drugs.
- AI role: An agent searched literature for crizotinib and lorlatinib analogues, curated 568 compounds, ran matched molecular pair analysis, benchmarked models, and proposed macrocycles and synthesis routes.
Evidence sources: https://deepmind.google/blog/agentic-cheminformatics-with-google-antigravity/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Vesuvius Challenge awarded a $50K First Letters Prize for scroll PHerc. 306B, where the winning submission reached the 10-character target and the papyrology team agreed on a reading, tentatively a Greek phrase with parallels in Aristotle and Philodemus. Ink was detected by multiplying two models (TRACE v3B and a fine-tuned A-hn). Letters remain faint, the readable area is small, and the work relied on hundreds of hours of manual labeling.
Technology: ML ink-detection models (TRACE v3B, A-hn) with VC3D GrowPatch
Publication: Announcement only · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: prospective or independent check
- Tested: The reading was checked by the papyrology team, which agreed on the letters with the winning submission, and was compared against 25 community submissions and a 10-character target; the page describes no independent replication.
- Still unknown: The reading is partial and uncertain, letters remain faint, training on machine-generated labels could compound errors (not assessed), and the Grand Prize is not yet awarded.
- AI role: Machine-learning ink-detection models (TRACE v3B and a fine-tuned A-hn) produced the ink maps; the readings and prize review were done by humans.
Evidence sources: https://scrollprize.substack.com/p/writing-seen-in-pherc-306b-for-the
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 4/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
In a prospective UK Biobank cohort of 27,429 participants without baseline chronic kidney disease, the study evaluated 2,920 plasma proteins and found 205 associated with incident CKD after protein-wide Cox regression with Bonferroni correction. LASSO reduced these to 23 candidates and LightGBM feature selection identified EDA2R, GDF15, HAVCR1 and SCARB2 as a four-protein panel. The protein panel alone had an AUC of 0.796 versus 0.805 for demographic variables; adding the proteins to the demographic model raised AUC to 0.830 in the study data.
Technology: Least absolute shrinkage and selection operator (LASSO) and Light Gradient Boosting Machine (LightGBM) feature selection
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: In a prospective UK Biobank cohort of 27,429 participants without baseline chronic kidney disease, the study evaluated 2,920 plasma proteins and found 205 associated with incident CKD after protein-wide Cox regression with Bonferroni correction. Protein selection and prediction were assessed internally within UK Biobank; prospective cohort collection does not establish prospective evaluation of a previously locked predictor.
- Still unknown: The abstract reports a prospective UK Biobank cohort and internal analyses; the four-protein model did not outperform the demographic model alone (AUC 0.796 vs 0.805, p=0.14), though combining the proteins with demographics improved AUC to 0.830. The authors state external validation is needed before clinical application. Protein associations and model selection do not establish causal roles or clinical utility.
- AI role: After protein-wide Cox association screening, LASSO selected 23 candidate proteins and LightGBM-based feature selection identified four proteins for the incident-CKD prediction panel. The study assessed their predictive contribution with demographic covariates and sensitivity analyses; these algorithms selected risk markers rather than demonstrating causal protein effects.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/42968231/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A Chain of Symbolic Regression (CoSR) analysis of hypersonic simulation data derived an interpretable non-power-product scaling relation for axial-force coefficient across sharp-cone and blunt-body configurations. It combines angle of attack, Mach number and Reynolds number through a dominant term and a viscous correction; the relation has a shared functional form but geometry-specific fitted constants. The authors interpret the constants as reflecting differences in shock/base drag, friction drag and pressure drag. This is a new model-derived scaling result from simulations, not a wind-tunnel or flight measurement.
Technology: Chain of Symbolic Regression (CoSR), a progressive symbolic-regression framework for dimensionless physical relations
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A Chain of Symbolic Regression (CoSR) analysis of hypersonic simulation data derived an interpretable non-power-product scaling relation for axial-force coefficient across sharp-cone and blunt-body configurations.
- Still unknown: The aerodynamic relation was inferred from hypersonic simulation data for two configurations (sharp cone and blunt body), not independently measured flight or wind-tunnel data. It is a parameterized model over the analyzed configuration/data range; external validation and broader geometry coverage are not established in the source. The simpler dominant component is linked to known hypersonic invariance, while the learned viscous correction and geometry-specific scaling provide the paper’s new modeled relationship. Do not treat the authors’ drag-component interpretations as direct measurements.
- AI role: CoSR progressively searched for interpretable dimensionless groups and combined their relationships instead of fitting one end-to-end expression. Applied to hypersonic simulation data, it identified a dominant angle-of-attack/Mach term and a Reynolds-number-dependent viscous correction, then fitted geometry-specific constants for two body shapes. Earlier canonical-law and fluid examples in the same paper are framework validations and are not counted separately here.
Evidence sources: https://journals.aps.org/prresearch/abstract/10.1103/f2kp-pp4q
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The preprint abstract reports computational generation of 500,000 candidate transcriptional activation domains and selection of 18,750 high-scoring sequences for yeast experiments using sequence-to-function models including three neural networks. More than 90% of 16,770 well-measured sequences functioned as activation domains. Some designed sequences retained activity in fungi, plants and human cells. Model agreement motivated the MADRIGAL consensus predictor. Review is limited to the abstract and original-server metadata.
Technology: Neural-network sequence-to-function models (three networks); consensus predictor MADRIGAL
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: 18,750 selected sequences were tested experimentally in yeast; over 90% of 16,770 well-measured sequences functioned as activation domains, and some retained function in fungi, plants and human cells.
- Still unknown: Review limited to the indexed abstract and publication metadata; full-paper methods and additional limitations were not assessed.
- AI role: Neural networks scored sequences generated from known activation-domain-like sequences; high-scoring sequences were selected for experimental validation and the consensus model MADRIGAL was developed.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.64898%2F2026.10.30.763198%22&resultType=core&format=json · https://api.biorxiv.org/details/biorxiv/10.64898/2026.10.30.763198
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Researchers measured a training library of yeast alcohol dehydrogenase I variants, used support-vector activity regression/classification and a thermostability model to rank combinations, then tested 22 prioritized variants. Four showed higher specific activity toward trans-2-nonenal than wild type; LWFMY also showed improved thermostability under the study’s tests. The result is laboratory enzyme engineering, not an industrial process or validated product. The primary journal version was published 2026-11-01; a search result for an earlier-hosted copy had ambiguous date/provenance, so this entry does not claim it was the first public version.
Technology: Support-vector regression/classification models and docking-informed mutation selection
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The top 22 ranked variants were experimentally evaluated; four outperformed wild type on trans-2-nonenal, with one also improved in thermal stability.
- Still unknown: Small selected validation set; specific activity differs from catalytic efficiency across substrate concentrations; ambiguous earlier-hosted copy means first-public date is not asserted.
- AI role: Models trained on measured variant data ranked combinations in unexplored sequence space for experimental testing.
Evidence sources: https://link.springer.com/article/10.1186/s13068-026-02839-2 · https://pubmed.ncbi.nlm.nih.gov/42972448/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 16 Nov 2026. Review scope: Primary Springer full article and PubMed metadata; methods, dates, affiliations and outcomes..
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Using 18 fb⁻¹ of Run 3 data and earlier LHCb data, the collaboration reports first single-experiment evidence for the rare B⁰→μ⁺μ⁻ decay, with a combined significance of 3.2σ and measured branching fraction (1.14 +0.42/−0.39 stat. +0.06/−0.05 syst.)×10⁻¹⁰. A boosted decision tree classified candidates and suppressed combinatorial background in the fit. This is evidence below the 5σ discovery threshold; results agree with Standard Model expectations. The date is the original preprint submission; CERN announced the result on 3 November 2026.
Technology: LHCb boosted decision tree (BDT) candidate classifier and multivariate trigger classifiers
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: Using 18 fb⁻¹ of Run 3 data and earlier LHCb data, the collaboration reports first single-experiment evidence for the rare B⁰→μ⁺μ⁻ decay, with a combined significance of 3.2σ and measured branching fraction (1.14 +0.42/−0.39 stat.
- Still unknown: This is a preprint submitted to Physical Review X and reports 3.2σ evidence, not a 5σ observation. Earlier joint LHCb–CMS evidence was not confirmed by later Run 1–2 analyses; this result is the first single-experiment evidence, not the first ever claim. The paper reports all measurements as consistent with Standard Model expectations. Machine learning contributes to candidate selection/background control along with substantial detector upgrades.
- AI role: The primary analysis uses a BDT to suppress combinatorial background, scores candidates and divides the dimuon mass fit into five response bins to improve signal sensitivity; it also uses multivariate classifiers in the software trigger. CERN states that machine-learning background suppression, together with detector upgrades, enabled the analysis to isolate these rare signals. The measured evidence comes from the collaboration's collision data and statistical fit, not an AI prediction.
Evidence sources: https://home.cern/lhcb-finds-evidence-of-one-in-ten-billion-particle-decay/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Arc Institute reports that Evo 2 designs functional regulatory sequences from 4 to 64 in-context examples without extra training, with 30-50% yield. Of nine AI-designed enhancers delivered by AAV to mice, seven drove detectable expression and two labeled layer 5 extratelencephalic neurons with 73% and 95% specificity. Results are bioRxiv preprints; in vivo work covers only layer 5 enhancers, and other designs (proteins, CRISPR loci, antibodies, phage genomes) are computational only.
Technology: Evo 2 (genomic language model); in-context learning; reinforcement learning post-training
Family Evo: profile · models page
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: Nine AI-designed enhancers were delivered by AAV to mice: seven drove detectable expression and two labeled layer 5 extratelencephalic neurons with 73% and 95% specificity.
- Still unknown: Only layer 5 enhancers were tested in vivo; the authors plan other cell types, and designs beyond enhancers (proteins, CRISPR loci, antibodies, phage genomes) are computational only.
- AI role: Evo 2 generated regulatory sequences from 4 to 64 in-context examples; reinforcement learning with predictive reward models sharpened cell-type specificity.
Evidence sources: https://arcinstitute.org/news/evo-2-in-context-learning
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Mutually reviewing LLM agents selected a question about the minimum number of calibration settings needed to predict responses to microwave pulse sequences in diamond spin ensembles, derived the theory and tested it experimentally. The first measurement was inconclusive because of signal drift, and after the protocol was revised the result matched the predictions under unchanged acceptance criteria. The abstract reports no numerical values, and this is an unrefereed arXiv preprint with an authors-reported validation.
Technology: Collaborative LLM agents that review each other's work
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: The agents derived the calibration rule computationally and ran diamond spin-ensemble experiments designed to distinguish predictions; the revised measurement matched the predictions under unchanged acceptance criteria.
- Still unknown: The abstract states no explicit limitations; the main caveat reported is signal drift in the first, inconclusive experiment.
- AI role: Mutually reviewing LLM agents generated and rejected research questions, derived theory, and designed and revised experiments.
Evidence sources: https://arxiv.org/abs/2610.40589
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The paper proves that every unital surjective complex-linear spectral isometry between unital C*-algebras is a Jordan isomorphism; without unitality, such a map is a central-unitary multiple of a Jordan isomorphism. The proof establishes that entire one-parameter groups of unital spectral isometries are algebra automorphisms, then reduces the general case to the commutative case. The authors report that the result was formally verified in Lean.
Technology: OpenAI GPT-6 Astra
Family GPT-6: profile · models page
Publication: Preprint · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: prospective or independent check
- Tested: The paper proves that every unital surjective complex-linear spectral isometry between unital C*-algebras is a Jordan isomorphism; without unitality, such a map is a central-unitary multiple of a Jordan isomorphism.
- Still unknown: ArXiv preprint. The authors explicitly describe the article as a human exposition/digestion of an AI-found proof. Their manuscript says formal verification is in Lean; The reported formalization was not independently audited in the source review.
- AI role: Authors state GPT-6 Astra found the proof after a sequence of simple prompts. They then discussed it with ChatGPT 6 to understand the mechanisms, wrote the paper themselves, and used ChatGPT 6 only for minor language editing; the Lean formalization is evidence for proof checking.
Evidence sources: https://arxiv.org/abs/2610.40470
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 4/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A Dartmouth preprint studied which language-model features predict human neural responses during naturalistic listening. Across three intracranial/fMRI datasets and six model conditions, reduced-rank regression recovered full linear encoding performance with about 1–3% of embedding dimensions. These directions differed from the embeddings’ principal variance axes and transferred across held-out participants, less strongly across datasets. This summary is limited to the abstract and affiliation/date metadata of the October 29, 2026 preprint. It concerns measured brain–model alignment, not proof that brains use the same representations.
Technology: Autoregressive and masked language models with ridge reduced-rank regression
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: Linear encoding of observed intracranial/fMRI responses, reduced-rank analysis and transfer across held-out participants.
- Still unknown: Source checking does not establish independent confirmation. Review date: 2026-11-15.
- AI role: LLM representations are the predictors of neural activity; reduced-rank regression identifies the brain-predictive subspace.
Evidence sources: https://arxiv.org/abs/2610.40511 · https://arxiv.org/html/2610.40511v1
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The journal abstract reports that machine-learning-assisted database mining identified 33 previously synthesized but unrecognized MXene candidates. Guided by that search, the researchers synthesized seven multilayer MXenes and realized 11 previously unexplored rare-earth-based MXenes. Experiments and modeling suggested semiconducting behavior and diverse magnetic states. Review is limited to the abstract and publication/affiliation metadata.
Technology: Machine-learning-assisted database mining for MXene candidates
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Seven multilayer MXenes were synthesized and 11 rare-earth MXenes realized; experiments and modeling assess electronic and magnetic behavior.
- Still unknown: Review limited to the indexed abstract and publication metadata; full-paper methods and additional limitations were not assessed.
- AI role: Machine learning performed database mining to screen candidate MXene compositions.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.1126%2Fscience.aeh5957%22&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Assay-aware models were used to select 84 new compounds for synthesis and prospective testing in a ROS1 kinase program, with TRKB as an anti-target. The most selective prospective compound had a measured ROS1 IC50 of 14.7 nM and 233-fold selectivity over TRKB, compared with 20.3 nM and 79-fold for the most selective training-set compound. The evidence is biochemical and does not establish therapeutic efficacy or safety.
Technology: D2ML multi-stage deep neural networks trained across direct-to-biology assay data
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: The authors prospectively synthesized and assayed 84 model-selected compounds. The most selective new compound had ROS1 IC50 14.7 nM and 233-fold selectivity over TRKB; the most selective training compound had 20.3 nM ROS1 IC50 and 79-fold selectivity.
- Still unknown: This is an in-vitro biochemical discovery result. It does not establish cellular activity, therapeutic efficacy, safety or clinical benefit. The paper compares a historical analog but makes no claim of global chemical novelty.
- AI role: Assay-aware models using data from multiple direct-to-biology screening stages selected compounds for prospective synthesis and ROS1/TRKB testing.
Evidence sources: https://arxiv.org/abs/2610.41070 · https://arxiv.org/html/2610.41070v1
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary arXiv v1 abstract, submission history and full text for date, affiliation, model methods and prospective assay result..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Anthropic reports early results with six partners: a CAR T-cell receptor design about 70% shorter that enabled immune cells to produce about 10 times more receptors; about 35,000 antibody designs in one month, yielding 20 candidates; and a document review that found all 11 human-identified issues plus seven more in 89% less time. The work is mostly preclinical and reported by Anthropic.
Technology: Claude
Family Claude: profile · models page
Publication: Announcement only · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: direct test
- Tested: Partner lab tests and a document review reported by Anthropic: a CAR T receptor design was tested in immune cells, antibody designs were screened for IgE binding, and a review of 1,061 pages was compared with 11 human-identified issues.
- Still unknown: The work is mostly preclinical with no human tests; partner reports give limited methodological detail, and the page does not say whether the seven extra flags were confirmed.
- AI role: Claude proposed molecular and receptor designs, generated about 35,000 antibody designs, reviewed manufacturing documents, and drafted regulatory documents.
Evidence sources: https://www.anthropic.com/research/early-progress-drug-development
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The paper reports a computational mechanism for perrhenate-catalysed deoxydehydration: proton-assisted water loss enables ring closure and alcohol reduction. Under human-defined hypotheses, a Claude Code agent located transition states, which the authors validated. Coupled-cluster-anchored microkinetic modeling predicted a rate increase of about eight orders of magnitude over the best proton-free pathway. At matched conversion conditions, modeled barriers were consistent with the 31.5–34.0 kcal/mol window inferred from previously reported experimental turnover; this was not a new wet-lab demonstration.
Technology: Claude Code LLM agent for human-directed reaction-network and transition-state search, combined with electronic-structure calculations and microkinetic modeling
Family Claude: profile · models page
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Agent-located transition states verified by the authors; computed energetics, reaction-network and microkinetic consistency checks; comparisons to existing turnover-derived barrier estimates.
- Still unknown: Main paper reviewed; supplementary files were not read. The mechanism and rate changes are predictions compared with prior experimental kinetics, not a new experimental confirmation.
- AI role: Claude Code carried out calculation/search/triage workflows for author-specified mechanistic hypotheses; authors validated the located stationary points.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.1039%2Fd6sc06865k%22&resultType=core&format=json · https://www.ebi.ac.uk/europepmc/webservices/rest/PMC13720772/fullTextXML
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
OpenAI announces results on local smoothing, a disproof of the exponential time hypothesis, and the construction of quantum field theories. The source presents these as research results to be understood with academic groups: physical artifacts are not yet public, and independent verification is not established here.
Technology: OpenAI internal models
Publication: Announcement only · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: prospective or independent check
- Tested: The local-smoothing proof was formalized in Lean the day after the AI output; the refutation of the Exponential Time Hypothesis was verified in Lean and is being checked by an external group led by Impagliazzo, Paturi and Williams.
- Still unknown: The physics result was not released; the external groups had not reported conclusions on the page.
- AI role: OpenAI internal models produced the proofs; external experts were asked to shepherd and explain the results.
Evidence sources: https://openai.com/index/building-scientific-understanding-with-academic-communities/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 5/5 · Validation 4/5 · Novelty 5/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A workflow combining a variational-autoencoder latent space of 11,069 literature photoredox reactions with random-forest catalyst/reagent recommendations was used to propose three transformations. Experiments synthesized a previously inaccessible substituted dihydroquinolinone 43 using fac-Ir(ppy)3, a related oxidative cross-coupling product 46 in 64% yield, and heterocycle product 49 in 74% yield. The paper presents these as experimentally validated transformations identified through model-assisted precedent retrieval and condition selection.
Technology: Variational autoencoder for reaction-space navigation and random-forest multilabel catalyst/reagent recommendation
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Experiments synthesized a previously inaccessible substituted dihydroquinolinone 43 using fac-Ir(ppy)3, a related oxidative cross-coupling product 46 in 64% yield, and heterocycle product 49 in 74% yield.
- Still unknown: The work is a human-guided photoredox design workflow, not autonomous synthesis. The three examples were selected by researchers and use a literature-trained model; the paper establishes feasibility for these tested transformations, not general transfer or broad yield improvement. Reported 64% and 74% yields apply only to the specific demonstrated products; further optimization and substrate-scope work are needed.
- AI role: The VAE organized 11,069 literature-derived reactions to retrieve traceable precedents; random-forest models recommended catalyst and reagent classes for proposed reactant-product pairs. Chemists used the model-suggested conditions to test three transformations, including two modifications beyond the reported precedent and one previously undocumented photoredox synthesis, then measured isolated product yields.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/42953598/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Iambic reported updated results from an open-label Phase 1/1b study of its AI-designed HER2 inhibitor IAM1363. At the August 31, 2026 cutoff, 122 participants had received treatment across four study parts; the company reported antitumor and intracranial responses and that more than half of participants treated at 960 mg remained on treatment. This summary is limited to the company release of October 26, describing an October 23 conference poster. These uncontrolled early results do not establish comparative efficacy.
Technology: Iambic molecular superintelligence platform
Publication: Announcement only · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: direct test
- Tested: Company-reported safety and preliminary response outcomes from an ongoing open-label Phase 1/1b trial.
- Still unknown: Source checking does not establish independent confirmation. Review date: 2026-11-15.
- AI role: The molecule was 'designed using Iambic's molecular superintelligence platform' and progressed to clinic in about two years, as stated by the company.
Evidence sources: https://ir.iambic.ai/news-releases/news-release-details/iambic-presents-updated-iam1363-clinical-data-esmo-2026-showing
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The October 25, 2026 paper applies depth-aware machine learning to copper-coated polymer-fiber mats and compares reconstructions with independent physical measurements. The abstract supports improved morphology agreement, but identifies no separate electrode-property or materials discovery beyond validating reconstruction quality. This source check is limited to the abstract, electronic publication date and affiliations; the full paper was not assessed.
Technology: Depth-aware machine learning: monocular depth estimation plus unsupervised clustering
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Reconstruction agreement with porometry, pycnometry and X-ray tomography for one fiber-mat material system.
- Still unknown: Admission review pending: primary full results needed to distinguish a specific scientific/application result from general method validation. Existing supported claims may be retained with bounded scope, but do not mark source checked until admission is resolved.
- AI role: Monocular depth estimation model, trained on macroscale images, plus unsupervised clustering to separate in-plane from subsurface material in FIB-SEM slices.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=EXT_ID%3A42928641&resultType=core&format=json
Needs source check · editorial details
The source has not been fully checked or contains conflicting information. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Nexus AI says it has deployed Google Research's chest X-ray AI as a certified diagnostic tool at about 40 facilities in six countries: more than 25,000 X-rays assessed by over 90 clinicians, with response times under 45 seconds. These are author-reported deployment figures, with no sensitivity or patient outcome data; WHO and FDA approvals are still in progress.
Technology: Foundation models for chest X-rays
Publication: Announcement only · Test context: Test context unspecified in annotation. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: not reported
- Tested: The post reports deployment figures (about 40 facilities, over 25,000 X-rays, over 90 clinicians) without sensitivity, specificity or patient outcome data from this deployment; it cites earlier peer-reviewed studies for performance.
- Still unknown: Deployment-level accuracy and outcomes are not given, WHO prequalification and FDA clearance are not granted, and the facility count across countries is unclear.
- AI role: The AI analyzed chest X-rays and flagged TB-suggestive findings for triage in under 45 seconds; clinicians made final decisions.
Evidence sources: https://research.google/blog/bringing-foundational-models-to-the-front-lines-of-tb-screening/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 1/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
PACT used healthy neural priors to decode hand gestures from recordings of patients with gliomas. In the reported comparison, decoding accuracy improved by more than 10%, and alignment with reference motor representations increased from r=0.66 to r=0.94. The study collected intraoperative ECoG from eight participants, four with motor-cortex gliomas and four with other gliomas. This proof of concept addresses a specific clinical population; offline decoding does not establish benefit from a deployed BCI. Review covered the abstract, affiliations and selected results.
Technology: Pathology-Aware Consensus Training (PACT), a neural decoder trained with biophysical constraints
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Offline gesture decoding on intraoperative recordings; selected results describe eight glioma participants.
- Still unknown: Source checking does not establish independent confirmation. Review date: 2026-11-15.
- AI role: PACT is the decoding model; it uses healthy neural recordings as priors and learns to suppress pathological noise and reweight spatial attention.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13693691/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The journal abstract describes CochleaNet, a deep-learning framework for cleared-cochlea light-sheet images, covering hair cells, spiral ganglion neurons, synapses, tonotopy and gene expression. Trained on mouse data, it was applied to gerbil and lower-resolution mouse and marmoset data. The authors say they analyzed neuron types and validated gene-therapy products but the abstract does not identify the biological or therapy-comparison findings. The date represents the 2026-10-23 journal version; an earlier preprint appeared on 2025-11-17.
Technology: CochleaNet: deep learning framework for volumetric image segmentation and analysis
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Manual-analysis and prior-segmentation comparisons; applications to neuron types, gene-therapy products, gerbil cochleae and lower-resolution mouse and marmoset imaging. The abstract does not state the specific biological outcomes.
- Still unknown: The journal body was not accessible through the publisher (HTTP 403), and the linked preprint full-text API returned HTTP 500. Original preprint abstract adds no specific gene-therapy result. Eligibility remains unresolved until the actual application finding is read.
- AI role: Deep learning framework that reconstructs the cochlea and segments hair cells, neurons and synapses, and quantifies tonotopy and gene expression.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.1016%2Fj.ymthe.2026.10.025%22&resultType=core&format=json · https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=EXT_ID%3APPR1120449&resultType=core&format=json
Needs source check · editorial details
The source has not been fully checked or contains conflicting information. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A 10-meter map classifies each land area into six forest or land-use types (including primary forest, plantations, and tree crops), with an initial time series from 2017–2025. Validation across more than 9,000 sites identifies the correct type in 90% of cases and distinguishes forest from non-forest in 95%.
Technology: Satellite-image classification models
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: The map was validated at more than 9,000 random sites, where two trained experts checked high-resolution imagery to determine land type: correct forest type in 90% of cases and forest vs non-forest in 95%.
- Still unknown: Per-class accuracy and regional breakdowns are not reported, the time series is early-release without separate validation, and results depend on the probability threshold chosen.
- AI role: A custom remote-sensing model trained on about 1.7 million labeled locations, using AlphaEarth embeddings, outputs probabilities for six forest and land-use classes.
Evidence sources: https://deepmind.google/blog/beyond-tree-cover-mapping-the-worlds-forest-types/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Machine-learning screening of LKB1-pathway genes in GEO data prioritized SIK1. In a PTZ-induced rat epilepsy model, SIK1 knockdown delayed seizures and shortened their duration, whereas overexpression worsened severity. The paper abstract also reports that restoring downstream NR4A2 counteracted SIK1-related effects in rescue and neuronal experiments. These are preclinical findings, not evidence of treatment efficacy in people. This summary is limited to the abstract and publication metadata.
Technology: Machine learning screening of LKB1 pathway genes on GEO dataset GSE143272
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Gene prioritization from GEO data followed by rat SIK1 perturbations, NR4A2 rescue and neuronal electrophysiology.
- Still unknown: Source checking does not establish independent confirmation. Review date: 2026-11-15.
- AI role: Machine learning screened LKB1 pathway genes and ranked SIK1 as the top epilepsy risk gene for experimental follow-up.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=EXT_ID%3A42991733&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
In the journal abstract, a language-model-inspired generative model trained on about 27,000 reef-fish surveys produced alternative coral-associated and macroalgal-associated community configurations under identical environmental conditions. It also inferred core, bridging and satellite species positions and improved occurrence and biomass predictions over XGBoost and MLP baselines. The configurations are model-derived and consistent with previously described ecological regimes; independent field confirmation is not established by the abstract.
Technology: Language-model-inspired generative model for ecological communities
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Fitted on about 27,000 reef fish surveys; predictive skill compared with XGBoost and MLP baselines; multistability interpreted against described ecological regimes.
- Still unknown: Review limited to the indexed abstract and publication metadata; full-paper methods and additional limitations were not assessed.
- AI role: Generative language-model-style model that predicts each species and its biomass conditional on previously assembled species, used to generate community configurations.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.1016%2Fj.cub.2026.09.081%22&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The Virtual Biotech journal abstract reports a human-guided multi-agent analysis of 55,984 trials. Drugs targeting cell-type-specific genes were associated with a 48% greater likelihood of reaching market and 32% fewer adverse events. The system also proposed a lung-cancer strategy and inferred possible reasons for a terminated ulcerative-colitis trial. These are retrospective associations and hypotheses, not demonstrated treatment benefits. The date represents the 2026-10-22 journal version; the linked preprint, including the same reported association, first appeared on 2026-02-23.
Technology: Virtual Biotech multi-agent AI framework (more than 37,000 agents)
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Retrospective trial-outcome/target-feature analysis across 55,984 trials, plus computational therapeutic and trial-failure hypotheses.
- Still unknown: Review limited to the indexed abstract and publication metadata; full-paper methods and additional limitations were not assessed.
- AI role: More than 37,000 AI agents annotated trial outcomes and integrated multimodal evidence to propose and analyze therapeutic decisions.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.1126%2Fscience.aeg6779%22&resultType=core&format=json · https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=EXT_ID%3APPR1222685&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The medRxiv abstract reports APOLLAI evaluation on video/audio recordings from 295 suspected-stroke patients at Nîmes University Hospital, against seven-neurologist consensus labels for five NIHSS items. Global sensitivity was 78.5% versus 59.4% for EMS personnel, specificity 87.0% versus 85.9%, and AUC 0.882. These are diagnostic comparisons on recorded patient data; the authors call for prospective validation in real EMS conditions.
Technology: APOLLAI (computer vision, audio deep learning and SVM fusion classifier)
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: Recorded cases from 295 suspected-stroke patients; AI and EMS assessments compared with consensus labels for five neurological tasks.
- Still unknown: Complete abstract and original-server metadata reviewed; full methods, model-selection timing and external EMS generalizability were not assessed. Authors explicitly call for real-world prospective validation.
- AI role: A multimodal machine-learning system detected stroke-related neurological deficits from video and audio recordings and was compared with clinicians' assessments.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.64898%2F2026.10.16.26365784%22&resultType=core&format=json · https://api.biorxiv.org/details/medrxiv/10.64898/2026.10.16.26365784
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Prospective study of 385 pregnant women in Chicago and Nairobi: the model estimates gestational age from scans taken by minimally trained operators with a mean error of 4.18 days, non-inferior to the 4.48 days of expert sonographers. This is a non-inferiority result from only two sites, both relatively urban.
Technology: Ultrasound model for estimating gestational age
Publication: Peer-reviewed paper · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: prospective or independent check
- Tested: A prospective multi-center study of 385 pregnant women at two sites (Chicago and Nairobi) compared AI gestational-age estimates from blind sweep scans by minimally trained operators with expert sonographer biometry: mean error 4.18 vs 4.48 days.
- Still unknown: The source does not state the non-inferiority margin or confidence intervals; only two relatively urban sites and one commercial probe were used, and clinical trust and confidence safeguards are not validated.
- AI role: The model estimated gestational age from blind sweep ultrasound scans and ran live on mobile devices in clinics.
Evidence sources: https://research.google/blog/empowering-frontline-maternal-care-with-ai-assisted-ultrasound/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 4/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Analysis of a CRISPR atlas of mouse brain cells identified differences between Tsc1 and Tsc2 and a possible relationship between DCTN1 and cholesterol regulation. Experts assessed the interpretations, but the proposed mechanisms remain hypotheses to be tested in new experiments.
Technology: GPT-Rosalind and distributed agents
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: Thousands of GPT-Rosalind agents reanalyzed an existing whole-brain CRISPR atlas; they reproduced findings from human analyses, and their novelty and value assessments aligned with blinded human experts.
- Still unknown: The proposed biological mechanisms still need experiments to test them.
- AI role: Each agent investigated one gene disruption, searched the literature and proposed explanations; reports were combined into a reasoning atlas.
Evidence sources: https://openai.com/index/scaling-ai-research-for-complex-biological-data/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The 16-billion-parameter Pinal model generated protein candidates from plain-language function descriptions. In four reported design tasks—fluorescent protein, PET hydrolase, alcohol dehydrogenase and metabolic H-protein—the selected designs displayed the intended function in laboratory tests; the best H-protein raised product titre 1.7-fold relative to native E. coli H-protein in a reconstituted CO2-fixation pathway.
Technology: Pinal, a 16-billion-parameter protein foundation model with text-to-structure (T2struct) and text-conditioned sequence design (SaProt-T) components
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: In four reported design tasks—fluorescent protein, PET hydrolase, alcohol dehydrogenase and metabolic H-protein—the selected designs displayed the intended function in laboratory tests; the best H-protein raised product titre 1.7-fold relative to native E. coli H-protein in a reconstituted CO2-fixation pathway.
- Still unknown: This is a primary-paper proof of concept with selected designs, not a broad hit-rate or industrial-performance demonstration. Experimental denominators vary substantially by task; extended data reports one active PETase among 94 screened and a fluorescent design selected from 29. The H-protein titre comparison was in a reconstituted pathway, not an organism or commercial process. No independent replication or therapeutic result is claimed.
- AI role: Pinal maps natural-language function descriptions to candidate protein structures and sequences. The team filtered and ranked model outputs, selected candidates for synthesis, and tested intended functions. The paper reports lab-observed fluorescence, PET-hydrolysis/esterase activity, alcohol-dehydrogenase catalysis and H-protein activity; the latter improved titre in a reconstituted multi-enzyme pathway.
Evidence sources: https://www.nature.com/articles/s41586-026-11205-z
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
In a mostly autonomous 64-hour campaign, Claude designed minibinders against five neurotoxins. External labs tested 211 designs, 17 of which met the criteria; the best (Cn2-B1 and Cn2-B2) had affinities in the low nanomolar range. Only recombinant toxins were tested, with no neutralization tests in cells or animals; Anthropic does not claim to have produced an antivenom.
Technology: Claude
Family Claude: profile · models page
Publication: Announcement only · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: direct test
- Tested: External labs tested 211 designs by surface plasmon resonance: Adaptyv Bio in cell-free expression and Twist Bioscience in human kidney cells; a binder needed KD of 1 µM or tighter in every replicate at both labs and no CLEC12A signal, giving 17 binders. The two collaborating external laboratories tested candidates from the discovery project; this establishes direct testing rather than a separately established independent follow-up replication.
- Still unknown: Only recombinant toxins were tested, with no neutralization in cells or animals; the authors do not know whether the binders neutralize venom and do not claim an antivenom; multitoxin designs failed.
- AI role: Claude selected toxin structures and hotspots, generated 318,484 designs, filtered and co-folded them, ran about 24 hours autonomously, then produced 180 selected designs after human follow-ups.
Evidence sources: https://www.anthropic.com/research/claude-scorpion-toxin-binder-design
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Bolero predicted cell-specific responses to AP-1 activity and transferred a human-trained response model to mouse and macaque neuronal enhancers, according to Arc Institute. Its announcement also describes a computational AP-1/cofactor spacing hypothesis; these virtual edits do not establish biological causality. This summary is limited to the October 20, 2026 institutional announcement, which identifies a collaboration with Harvard Medical School and describes a 10.8-million-cell atlas. The linked paper is a preprint; full-paper methods and experimental independence were not assessed.
Technology: Bolero (cell-state-conditioned sequence-to-function Transformer)
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: Human-trained AP-1 response predictions compared with mouse/macaque neuronal enhancers; virtual regulatory perturbations.
- Still unknown: Source checking does not establish independent confirmation. Review date: 2026-11-15.
- AI role: Bolero, a Transformer conditioned on cell-state embeddings, predicts cell-state-specific gene regulation directly from DNA and estimates responses in unprofiled cell states.
Evidence sources: https://arcinstitute.org/news/bolero
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
ProVADA steered a pretrained protein generator to produce active, drug-inhibitable cytosolic renin and deimmunized two murine antibody fragments while retaining antigen recognition. ESCAPE-seq measured reduced HLA-I peptide binding, and redesigned FMC63 variants retained CAR-T-cell killing. This summary is limited to the abstract of the October 18, 2026 version 2 preprint; the initial preprint dates to July 17, 2025.
Technology: ProVADA (Mixture-Adaptation Directed Annealing sampler steering a pretrained protein generator)
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: Renin activity/inhibitor response, antibody antigen recognition and HLA-I peptide binding, and FMC63 CAR-T-cell killing, as reported in the version 2 abstract.
- Still unknown: Only the abstract, PMC version header and original-server corresponding-institution metadata were assessed; full author-affiliation list and full paper were not reviewed. Preserve ID.
- AI role: A pretrained protein generator, steered by ProVADA's annealing sampler toward externally defined properties while preserving functional constraints.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13742580/ · https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.1101%2F2025.07.11.664238%22&resultType=core&format=json · https://api.biorxiv.org/details/biorxiv/10.1101/2025.07.11.664238
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
AI agents designed DNA aptamers and protein binders, which were then tested in the lab over multiple rounds. Target success rates reached 33.5% for myoglobin and 42.1% for α-thrombin (98.5% specificity). The post summarizes three arXiv papers whose peer review is unconfirmed; it also reports initial failures.
Technology: AI agents with laboratory robots
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: Agents ran three design-test-iterate rounds of DNA aptamers against myoglobin and alpha-thrombin in the lab, reaching 33.5% and 42.1% hit rates with 98.5% specificity for thrombin; protein binders against FAP were reported at single-digit nanomolar affinity.
- Still unknown: Peer review of the cited arXiv papers is unconfirmed; there are two protein targets and three rounds; initial rounds failed, and protein expression adds steps to each loop.
- AI role: Agents designed and revised sequences across rounds and adjusted design rules after reading wet-lab data, under human oversight of protocols and robot runs.
Evidence sources: https://research.google/blog/grounding-ai-agents-in-biological-reality-building-blocks-for-autonomous-wet-labs/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Claude designed circuits for eight basic operations (chemistry, optimization, physical simulation, and number theory), assessed by an automated verifier. In several cases they outperform the best published circuits; for example, a point addition used 17 times fewer gates at the cost of 12% more qubits. Key identities were checked with simulation and Lean; a memory shortcut in the decoder fails about once in 10,000 times.
Technology: Claude
Family Claude: profile · models page
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: An automated evaluator checked each circuit for correctness and scored qubits times gates; results were compared with published designs, divsteps was checked by simulation and Lean, and decoder memory was measured over 403 million simulated decodings.
- Still unknown: The decoder memory shortcut fails about once in 10,000 runs; gains apply to individual primitives, not a full fault-tolerant machine; Claude was conservative and needed prompts to broaden its search.
- AI role: Claude wrote circuits as Python code and revised them in a closed loop with evaluator feedback, with no human input between rounds, and reconstructed branch information to reduce memory.
Evidence sources: https://www.anthropic.com/research/designing-efficient-quantum-circuits
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
An agentic workflow searched UniProt for antimicrobial candidates in Streptococcus pseudopneumoniae and found a 97-amino-acid bacteriocin with no homologues among manually reviewed Swiss-Prot proteins. This is a computational candidate now undergoing laboratory validation; no potency or toxicity data are available.
Technology: Google Antigravity (agentic platform)
Publication: Announcement only · Test context: Test context unspecified in annotation. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: not reported
- Tested: Nothing was tested in the lab: the candidate's novelty was checked by BLAST against Swiss-Prot, its predicted structure was compared with a known bacteriocin fold, and its genomic context was examined; the wet-lab plan is underway with no results.
- Still unknown: No potency or toxicity results exist; the wet-lab assays outcomes are unknown; the session covered a narrow sequence window in one species.
- AI role: An agent queried databases, filtered candidates, ran BLAST and structure comparisons, analyzed genomic neighborhoods, synthesized literature and drafted a wet-lab plan under researcher direction.
Evidence sources: https://deepmind.google/blog/accelerating-antimicrobial-discovery/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 1/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Researchers trained a large-language-model framework to predict fluorophore emission wavelengths and used virtual screening to prioritize NIR-II imaging molecules. Two candidates were experimentally characterized and showed favorable imaging performance in vitro and in vivo, according to the abstract. The abstract supplies no quantitative imaging effect sizes. This summary is limited to the abstract and October 15, 2026 publication metadata; the candidates are imaging probes.
Technology: Large language model for structure-property prediction, with virtual screening
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Two computationally prioritized fluorescent candidates characterized for imaging in vitro and in vivo.
- Still unknown: Source checking does not establish independent confirmation. Review date: 2026-11-15.
- AI role: A large language model predicted emission wavelengths and guided virtual screening to choose candidates.
Evidence sources: https://onlinelibrary.wiley.com/doi/10.1002/smll.76251?af=R
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
IRIS inferred signaling histories from gene activity after training on human stem-cell perturbations. The researchers used it to identify a signaling intervention predicted to encourage lung development and report experimental confirmation in mouse embryos. This summary is limited to MIT/Whitehead’s October 15, 2026 announcement, which explicitly dates the Nature Methods paper to September 8. The release supplies no quantitative accuracy metrics; proposed organoid and therapeutic uses are future applications.
Technology: IRIS (neural network with transfer learning)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Inferred cell signaling histories and a lung-development prediction checked in mouse embryos, according to the institutional release.
- Still unknown: Source checking does not establish independent confirmation. Review date: 2026-11-15.
- AI role: IRIS learns consistent pathway fingerprints in gene activity across cell types, reconstructs signaling histories, and generated the lung-development prediction that was tested in mouse embryos.
Evidence sources: https://news.mit.edu/2026/ai-model-decodes-language-cells-use-communicate-1015
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
RFdiffusion2 for Molecular Interfaces designs zinc-dependent proteases for peptide-bond hydrolysis. The primary paper reports cleavage at the targeted bond in 36% of 135 designs and selected designs that cleave disease-associated proteins in vitro. These biochemical results do not establish therapeutic efficacy.
Technology: RFdiffusion2 for Molecular Interfaces (RFD2-MI); AlphaFold 3
Family RFdiffusion: profile · models page
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The source reports the study-specific result summarized here.
- Still unknown: Independent replication and broader deployment or clinical benefit are not established by this source review.
- AI role: RFD2-MI generates scaffolds around zinc catalytic and peptide-binding geometries; computational filtering guided experimental testing.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC12667880/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 16 Nov 2026. Review scope: Primary-source abstract and publication metadata reviewed; full text scope is specified in source_checks..
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 5/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
For the Mw 7.4 Drake Passage earthquake of 2 May 2025, a trained classifier identified depth-sensitive pP–sP seismic phases used to refine focal depths under sparse station coverage. The authors infer a mainshock depth of about 25 km, consistent with the small tsunami, and report aftershocks on west-dipping faults southeast of the epicenter near the Shackleton Fracture Zone; the paper integrates this with seismic-reflection and previously collected imaging data.
Technology: Trained machine-learning classifier for depth-sensitive seismic phase identification
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: For the Mw 7.4 Drake Passage earthquake of 2 May 2025, a trained classifier identified depth-sensitive pP–sP seismic phases used to refine focal depths under sparse station coverage.
- Still unknown: The result is a retrospective analysis of one earthquake sequence. The depth is inferred from phase picks and is consistent with tsunami observations; tectonic interpretation also relies on aftershock mechanisms and independent seismic imaging. The machine-learning component contributes phase identification rather than discovering the earthquake itself.
- AI role: Identified pP–sP phases in regional records; these phases were incorporated into event relocation and focal-depth determination, helping constrain the mainshock depth where station coverage is limited.
Evidence sources: https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2026GL122178
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A two-phase dense-array catalog processed with the LOC-FLOW machine-learning workflow revealed microseismicity in three zones and shallow (0–6 km) and deep (6–15 km) layers near the northern Xiannvshan Fault Zone. A steeply dipping seismic band is interpreted as a previously hidden western extension of the Tianyangping Fault; the authors also propose fluid-permeable connections between depth layers. The enhanced catalog detects over an order of magnitude more events than the regional network catalog for the same period.
Technology: LOC-FLOW machine-learning seismic catalog workflow
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A two-phase dense-array catalog processed with the LOC-FLOW machine-learning workflow revealed microseismicity in three zones and shallow (0–6 km) and deep (6–15 km) layers near the northern Xiannvshan Fault Zone.
- Still unknown: Verified from the original publisher abstract and publication metadata; full text was inaccessible in this review. The steep band is interpreted as an inferred buried fault, not independently imaged or proven fault geometry. The proposed fluid-permeable connection and dual-mechanism reservoir-modulated seismicity are hypotheses. The catalog increase reflects a dense temporary array plus machine-learning workflow, compared with the regional network catalog.
- AI role: The study used LOC-FLOW with dense-array observations to detect, locate and relocate microearthquakes for a higher-resolution catalog. The catalog made the three activity zones, depth layers and steep seismic band visible for the authors’ structural interpretation.
Evidence sources: https://www.sciencedirect.com/science/article/pii/S1674984726001278
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The paper proves that a uniformly random permutation of length 32n² contains every permutation of length n as a pattern with probability at least 1−exp(−c n/log n), for an absolute c>0 and sufficiently large n. This removes the prior log-log factor from the sufficient-length bound, resolves Alon’s open 1000n² problem, and establishes the correct Θ(n²) order.
Technology: OpenAI GPT-6 Pro
Family GPT-6: profile · models page
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: The paper proves that a uniformly random permutation of length 32n² contains every permutation of length n as a pattern with probability at least 1−exp(−c n/log n), for an absolute c>0 and sufficiently large n.
- Still unknown: ArXiv preprint. Authors report checking and digesting the proof; no separate independent validation is claimed here. The result is an asymptotic sufficient bound with constant 32 (the paper explains the constant can be reduced to any value greater than 1), not a sharp threshold constant.
- AI role: The authors state that GPT-6 Pro found the proof; they do not claim credit for its proof ideas, and say they verified and digested it and take responsibility for the manuscript.
Evidence sources: https://arxiv.org/abs/2610.13530
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A genome-wide imaging and CRISPR screen of about 70 million cells identified GPR89A and GPR89B (encoding GPHR) as important for maintaining Golgi acidity and early STING trafficking and activation. The study also found a previously undescribed autophagy phenotype associated with EI24 and linked it to altered autophagosome formation.
Technology: SPARCS platform with machine-learning image analysis
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: A genome-wide imaging and CRISPR screen of about 70 million cells identified GPR89A and GPR89B (encoding GPHR) as important for maintaining Golgi acidity and early STING trafficking and activation.
- Still unknown: The study used pools of at least 500 excised cells for sequencing, so it could not map individual images to individual guide RNAs. The paper describes GPHR as a regulator of STING trafficking and signaling; therapeutic implications remain untested.
- AI role: Machine-learning image analysis classified phenotypes in genome-wide CRISPR screens; SPARCS recovered known autophagy genes and surfaced EI24 and GPHR-linked findings. Follow-up molecular profiling and experiments by the researchers examined these hits.
Evidence sources: https://www.helmholtz-munich.de/en/icb/news-detail/new-method-links-cell-images-to-genetic-changes
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Using 140 fb⁻¹ of 13 TeV proton–proton collisions, ATLAS reports the first observation of single-top-quark s-channel production at the LHC, with 5.1σ observed (5.2σ expected) significance. The measured cross-section is 10.8 ± 1.5(stat) ± 1.6(syst) pb, consistent with the Standard Model. A graph neural network discriminant separates the signal from top-pair, W+jets and t-channel single-top backgrounds; the analysis also uses broadened event acceptance, improved b tagging and refined modeling.
Technology: Event-level graph neural network; ATLAS GN2 graph-transformer b-jet tagger
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: Using 140 fb⁻¹ of 13 TeV proton–proton collisions, ATLAS reports the first observation of single-top-quark s-channel production at the LHC, with 5.1σ observed (5.2σ expected) significance.
- Still unknown: This is an ATLAS preprint submitted to PRL as of the checked date; prior ATLAS analyses had reported evidence, and the Tevatron had already observed the process in a combined CDF+DØ result. The new claim is first observation at the LHC and by one experiment. The cross-section has about 20% total uncertainty and agrees with the Standard Model; the network is one of several improvements contributing to sensitivity, not the sole cause.
- AI role: The GNN uses event kinematics to distinguish s-channel signal from major backgrounds, with its score distribution used in the likelihood fit. GN2 improves b-jet identification; the paper attributes increased sensitivity jointly to the GNN, broader acceptance, improved b tagging, calibrations and theory modeling.
Evidence sources: https://atlas.cern/Updates/Briefing/Run2-Top-s-Channel
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The PDFM model increased the explained variance in MMR vaccination coverage from 16% to 22% across 146 counties on the US–Canada border, and improved weekly correct cholera-risk-area rankings in the DRC from 1.78 to 2.10. Gains for cardiovascular mortality and postpartum depression were small or not significant.
Technology: Population Dynamics Foundation Model (PDFM), TimesFM
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: Five partner-led evaluations compared PDFM embeddings with baselines: MMR explained variance rose from 16% to 22% across 146 border counties and cholera Precision@5 improved from 1.78 to 2.10; cardiovascular gains were not significant.
- Still unknown: Several results were non-significant or small, cholera gains were not significant at one or two weeks, and the authors cite static embeddings as a limitation.
- AI role: PDFM combines search trends, mobility, built-environment and environmental data into location embeddings used without task fine-tuning; for dengue it was paired with TimesFM 2.0.
Evidence sources: https://research.google/blog/earth-ais-planetary-geospatial-foundation-models-for-global-public-health/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The source presents a proof discovered autonomously by GPT-6 Pro establishing an O_d(r^{-1/2}) zero-probability bound for degree-d multilinear polynomials with r disjoint degree-d monomials, improving the prior bound by removing a polylogarithmic factor and resolving the Nguyen–Vu conjecture. The key argument also proves a total-influence estimate for bounded-degree rational functions, resolving a recent conjecture of Kothari, Kovacs-Deak, Wang and Yang; these are one connected paper-level result, not separate entries.
Technology: OpenAI GPT-6 Pro
Family GPT-6: profile · models page
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: The source presents a proof discovered autonomously by GPT-6 Pro establishing an O_d(r^{-1/2}) zero-probability bound for degree-d multilinear polynomials with r disjoint degree-d monomials, improving the prior bound by removing a polylogarithmic factor and resolving the Nguyen–Vu conjecture.
- Still unknown: This is an arXiv preprint. The author reports personally verifying and rewriting the proof; this review has not independently formalized it. Bound and related rational-function influence result are counted once because the paper derives them as a connected argument.
- AI role: The author states that GPT-6 Pro autonomously discovered the proof and that the author claims no credit for the proof ideas; the author subsequently verified, digested and rewrote the argument.
Evidence sources: https://arxiv.org/abs/2610.08708
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Nathaniel Craig proposes a smooth, exponentially local, gauge-invariant fermion measure on admissible four-dimensional lattices, with restrictions covering the Standard Model gauge groups. GPT-6 Astra helped develop the refinement construction and led the detailed proof. The supplementary report states that Lean kernel replay accepted the encoded measure-existence result and supporting inputs under their recorded assumptions. This is a candidate finite-lattice definition; reflection positivity, practical simulation and an interacting continuum theory remain outside the established claim. The October 6, 2026 preprint preceded the November 15 institutional announcement.
Technology: GPT-6 Astra; Lean 4 formalization
Family GPT-6: profile · models page
Publication: Preprint · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: prospective or independent check
- Tested: Source-reported Lean kernel replay checked the concrete companion measure-existence theorem, its analytic/geometric inputs and Standard Model restrictions under encoded definitions and assumptions. Same-study mathematical and two-dimensional computational checks are also reported.
- Still unknown: Review checked the preprint verification/conclusion sections and formal-verification supplement, not an independently executed Lean build. Formal-to-source correspondence was reviewed by AI agents. A separate frozen-reference comparator failed on generated proof names; the report distinguishes this from successful kernel replay. Kernel correctness is relative to the encoded claim and axioms; it does not establish reflection positivity, a physical continuum limit, independent scientific confirmation or practical usefulness.
- AI role: Craig supplied the refinement strategy; Astra developed key construction steps, led proof development and carried out proposed verification work.
Evidence sources: https://arxiv.org/html/2610.09093v1 · https://arxiv.org/src/2610.09093v1/anc/lean-verification-report.pdf · https://arxiv.org/src/2610.09093v1/anc/mathematical-companion.pdf · https://openai.com/index/step-toward-defining-the-standard-model/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 4/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Alman and Vassilevska Williams present deterministic algorithms for 3SUM on polynomial-size integers in O(n^1.9992) time and all-pairs shortest paths on directed graphs with polynomially bounded integer weights in O(n^2.9995) time, the first polynomial improvements over the textbook exponents. The single underlying technique computes selected entries of thin matrix products more efficiently, and known reductions yield the main algorithmic consequences. The paper says an internal Anthropic Claude research model discovered the algorithm autonomously; the human authors then understood, simplified, strengthened and extended it. One paper-level theoretical-computer-science result is counted.
Technology: Internal Anthropic Claude research model (specific model version undisclosed)
Family Claude: profile · models page
Publication: Preprint · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: prospective or independent check
- Tested: Alman and Vassilevska Williams present deterministic algorithms for 3SUM on polynomial-size integers in O(n^1.9992) time and all-pairs shortest paths on directed graphs with polynomially bounded integer weights in O(n^2.9995) time, the first polynomial improvements over the textbook exponents.
- Still unknown: This is an arXiv preprint as of the checked record. The stated time bounds are for polynomially bounded integer inputs; real-valued variants use Las Vegas randomization and different bounds. Lean certification covers the explicitly enumerated theorems, not every statement or presentation detail in the 76-page paper. The authors take responsibility for the paper and developed important extensions/consequences after the initial AI-discovered algorithm.
- AI role: The paper states that Claude was tasked with checking and improving cryptographic constructions but instead developed the new thin-matrix-product algorithm, first for average-case and then worst-case problems, in a 16-million-output-token session with no human input. The authors subsequently worked to understand, simplify, strengthen and extend the algorithm. Anthropic later used an internal model to Lean-certify specified main results.
Evidence sources: https://arxiv.org/abs/2610.06783
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 4/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
In numerical simulations of a one-dimensional Landau–Ginzburg phase transition, a recurrent neural network predicted the final locations of topological defects from short sequences of order-parameter fluctuations during the Kibble–Zurek impulse regime, before freeze-out and before the defects were macroscopically visible. Prediction error across quench rates followed power-law scaling consistent with the Kibble–Zurek freeze-out time, indicating that information about the eventual defect pattern is encoded in near-critical fluctuations in this model.
Technology: Recurrent neural network (RNN); LSTM and Transformer architectures were also checked
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: In numerical simulations of a one-dimensional Landau–Ginzburg phase transition, a recurrent neural network predicted the final locations of topological defects from short sequences of order-parameter fluctuations during the Kibble–Zurek impulse regime, before freeze-out and before the defects were macroscopically visible.
- Still unknown: This is a computational result from a noisy one-dimensional Landau–Ginzburg model, not an experimental observation of defects in a material or cosmological system. The paper uses overdamped dynamics (η=1), low noise temperature θ=10⁻⁸, a finite periodic grid and simulated training/test samples; similar behavior at another noise setting is reported in supplementary material. The finding is consistent with the Kibble–Zurek mechanism and an earlier deconstruction of its dynamics; it does not establish that this prediction method applies unchanged to all phase transitions. The PRL article was published 2026-10-05; the original arXiv preprint was submitted 2025-08-28 and revised as v2 on 2026-09-10.
- AI role: The RNN analyzed short time sequences from numerical order-parameter dynamics and inferred the final kink count and positions from small fluctuations during the impulse regime. This allowed the authors to identify when information about future defect positions becomes predictable, a quantity not directly provided by conventional Kibble–Zurek defect-density scaling. They report similar prediction behavior with LSTM and Transformer models, suggesting the result is not specific to one architecture.
Evidence sources: https://journals.aps.org/prl/abstract/10.1103/fstv-lsvh
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Zhao Song and Xinzhi Zhang prove a universal, dimension-independent bound for the Kannan–Lovász–Simonovits constant of isotropic log-concave probability measures, resolving the KLS conjecture. Their proof repeatedly refines bounds on polynomial variance and curvature; the authors report that AI tools helped explore more than 100 approaches, obtain the earlier log-star bound, and develop and verify the final proof. Independent researchers also posted proofs, and the paper’s streamlined proof is being formalized in Lean. The 1 October preprint gave a slowly dimension-dependent bound; the 4 October version established the dimension-independent theorem described here.
Technology: GPT-6 Astra; GPT-5.6 Sol; Claude Fable 5; Fable (as named in the paper’s AI disclosure)
Family Claude: profile · models page
Publication: Preprint · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: prospective or independent check
- Tested: Zhao Song and Xinzhi Zhang prove a universal, dimension-independent bound for the Kannan–Lovász–Simonovits constant of isotropic log-concave probability measures, resolving the KLS conjecture.
- Still unknown: The proof is a recent arXiv preprint, not yet journal peer-reviewed in the source reviewed. Independent proofs were posted, and the Microsoft Research author account reports later Lean formalization of a streamlined proof; the paper version and posted revisions should be checked as the work develops. This entry counts the theorem result once, not the intermediate log-star bound separately.
- AI role: The authors state that they explored candidate approaches collaboratively with AI tools; the tools helped obtain the earlier log-star bound and develop detailed proof steps, while the researchers selected the direction and completed the repeated refinement argument. The authors state they checked the proofs with AI tools and take responsibility for correctness.
Evidence sources: https://arxiv.org/pdf/2610.01447v2 · https://arxiv.org/abs/2610.01447 · https://www.microsoft.com/en-us/research/blog/solving-a-31-year-old-grand-challenge-in-mathematics-the-kls-conjecture/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 4/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The authors present a proof of the Linear Hadwiger conjecture: there is a constant C such that every graph with no K_t minor is colorable with at most Ct colors. GPT-6 Astra developed the proof after the authors directed it to the dense case and then asked it to bridge existing bounds; the resulting proof combines several known frameworks with a new bootstrap argument. The result is an asymptotic linear bound, not the stronger Hadwiger conjecture χ(G)≤h(G).
Technology: OpenAI GPT-6 Astra (described in the paper as ChatGPT model Astra 6)
Family GPT-6: profile · models page
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: The authors present a proof of the Linear Hadwiger conjecture: there is a constant C such that every graph with no K_t minor is colorable with at most Ct colors.
- Still unknown: The result is an arXiv preprint and is not the full Hadwiger conjecture. It proves existence of an absolute linear-coloring constant for K_t-minor-free graphs. The authors take responsibility for the proof; Independent formal verification was not assessed in the source review.
- AI role: Authors disclose that the proof was developed by the model following their research directions and feedback; the model supplied a new bootstrap argument after being prompted first on a dense case and then to bridge existing bounds. Authors independently wrote the exposition, reviewed the result and take responsibility for correctness.
Evidence sources: https://arxiv.org/abs/2610.05291
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
MetasynBGC used gene-cluster-family-guided metagenomic mining and a Chemprop graph neural network to prioritize predicted natural-product structures for synthesis. Six selected NRPS-derived compounds showed in-vitro cytotoxic activity across seven cancer cell lines, with reported IC50 values as low as 38.04 μM; compounds D and E showed cell-line-selective activity. These are in-vitro results and do not establish in-vivo efficacy or therapeutic benefit.
Technology: Chemprop graph neural network used to prioritize predicted biosynthetic products by antitumor activity
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Six synthesized NRPS-derived compounds were tested in vitro across seven cancer cell lines; the paper reports IC50 values as low as 38.04 μM and selective activity for compounds D and E.
- Still unknown: No in-vivo efficacy or clinical therapeutic benefit is established. The reported potency is limited to the tested in-vitro cancer cell-line assays.
- AI role: A Chemprop graph neural network scored PRISM-predicted biosynthetic product structures for antitumor potential; the authors jointly used predicted scores and structural feasibility to prioritize compounds for synthesis and cytotoxicity testing.
Evidence sources: https://journals.asm.org/doi/10.1128/spectrum.01098-26
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary ASM article abstract, methods and author affiliation block checked for online date, model role, synthesized compounds and measured cytotoxicity..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Methods insert a detectable signal into protein sequences and predicted structures. For binders designed against three targets (VEGF-A, SARS-CoV-2 RBD, and PD-L1), watermarked designs matched the others in success rate and affinity. This is a proof of concept that needs to be made more robust against tampering.
Technology: SynthID Bio, AlphaFold 3 (optimized component)
Family AlphaFold 3: profile · models page
Publication: Announcement only · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: direct test
- Tested: Wet-lab tests on three targets (VEGF-A, SARS-CoV-2 RBD, PD-L1), with Adaptyv Bio helping, showed watermarked designs matched unwatermarked ones in hit rate, affinity and diversity; the phage was tested only in bacterial cultures.
- Still unknown: The watermark must become more robust against deliberate tampering, the bacteriophage work is early-stage, and the post says no single biosecurity measure is a silver bullet.
- AI role: SynthID Bio guided amino acid choice in a ProteinMPNN variant, adjusted predicted atomic coordinates, fine-tuned part of AlphaFold 3, and was integrated into Evo 2 to watermark a designed phage genome.
Evidence sources: https://deepmind.google/blog/introducing-synthid-bio/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
WDNet detected candidates in DESI Legacy Surveys DR9 images. The researchers cross-matched 18,242 photometric candidates with DESI DR1 spectra, obtaining 719 candidate spectra; manual visual inspection of those spectra identified 75 new white dwarfs and 7 new cataclysmic variables. These spectroscopic classifications, not the much larger image-candidate list, are the reported verified sample.
Technology: WDNet, a deep-learning object-detection network based on YOLOv8 with multi-scale feature fusion
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: WDNet detected candidates in DESI Legacy Surveys DR9 images.
- Still unknown: The authors confirmed 75 new WDs and 7 new CVs among the 719 candidates with DESI spectra, and explicitly note substantial contamination among remaining candidates and selection effects in the spectroscopic subsample. No atmosphere-model fitting was done for all identifications; the authors recommend further high-S/N spectroscopic modeling. The article's abstract/results/conclusion give inconsistent counts for the full detected/candidate catalog (18,242, 61,624, and 63,214 are all reported in different sections); this entry relies only on the consistent 719 spectral matches and 75+7 inspected confirmations. The full photometric catalog is not counted as confirmed discoveries.
- AI role: WDNet detected and localized white-dwarf candidates in DESI Legacy Surveys photometric images, thereby prioritizing objects for DESI spectral cross-match and manual spectral inspection. Human inspection of DESI spectra supplied the classification confirmation; the neural network did not itself confirm the new stellar classes.
Evidence sources: https://www.mdpi.com/2218-1997/12/10/295
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Microsoft's Quine, a multimodal biology world model, prioritized thousands of compounds for shifting tumor cell states, and a handful were tested in pancreatic ductal adenocarcinoma cell lines. The top-ranked compounds produced the largest classical-to-basal shifts, as validated by wet-lab experiments, and some moved cells toward a distinct third phenotype. The post gives no quantitative wet-lab figures and links no paper, and it measures cell-state changes rather than clinical efficacy.
Technology: Quine, multimodal biology world model with interactive harness
Publication: Announcement only · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: direct test
- Tested: A handful of top-ranked compounds predicted by Quine were tested in patient-derived pancreatic ductal adenocarcinoma cell lines, where the largest classical-to-basal transcriptional shifts were reported.
- Still unknown: The page calls the drug-repurposing findings early evidence and does not report how many compounds were tested or the assay statistics.
- AI role: Quine, a multimodal biology world model with a harness, ranked thousands of compounds by their predicted ability to shift tumor cell states.
Evidence sources: https://microsoft.com/en-us/research/blog/introducing-quine-an-ai-research-system-designed-for-the-complexity-of-biology
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A PCA enrichment map and machine-learning virtual screening pipeline prioritized a commercial library for beta-haematin and parasite assays. The primary indexed paper reports several chemotypes evaluated for antiplasmodial activity, beta-haematin inhibition and haemozoin formation inhibition. This is an in-vitro campaign; results should not be conflated across assays or with clinical efficacy.
Technology: PCA enrichment mapping; support-vector machine classification; virtual screening
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: PMC-hosted primary text reports PCA/ML screen and experimental parasite/haemozoin assays.
- Still unknown: Exact structure-level overlaps with earlier UCT screens and detailed full-text quantitative results should be cross-checked before final prioritization.
- AI role: PCA map and classifier prioritize compounds for biochemical and cell assays.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/42857431/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: PMC-indexed primary article; PubMed primary metadata. Material checked for the specific claims stated in the record..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A generative deep neural network searched multilayer indium-tin-oxide/aluminum-doped-zinc-oxide designs and selected an epsilon-near-zero absorber stack. After the researchers merged two similar adjacent layers for fabrication, the approximately 159-nm device experimentally maintained more than 98% absorption over a 1,077-nm near-infrared bandwidth at 49° p-polarized illumination. The modeled design predicted 1,688 nm bandwidth; the experiment’s detector cutoff limited measurements above 2,600 nm.
Technology: GLONet generative deep neural network, trained in simulation loop with differentiable transfer-matrix solver TMM-Fast
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: A generative deep neural network searched multilayer indium-tin-oxide/aluminum-doped-zinc-oxide designs and selected an epsilon-near-zero absorber stack.
- Still unknown: The demonstrated result is a fabricated device under p-polarized illumination at a 49° incidence angle using a prism-coupled ENZ mode; it is not an angle-independent or application-qualified absorber. Measurement above 2,600 nm was limited by the detector, so the 1,688-nm bandwidth is a simulation result and the measured bandwidth is 1,077 nm. The network-produced four-layer design was manually simplified by merging two similar layers before fabrication. Earlier work by the same group used neural networks to design homogeneous ITO stacks; this paper reports a distinct heterogeneous ITO/AZO design with experimental validation. No commercial application is demonstrated.
- AI role: The GLONet maps random latent vectors to multilayer designs, optimizing per-layer refractive/material parameters and thickness using differentiable optical simulations to maximize broadband absorption. The authors generated 10,000 candidate stacks and selected a high-performing design for experimental fabrication and measurement. A manual layer-merging step simplified the four-layer design to three layers before fabrication.
Evidence sources: https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.77966
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
An AI-guided, human-supervised platform assessed 2,942 catalysts and identified complex palladium oxides. InMnPdOx maintained an overpotential below 0.5 V for more than 1,000 hours in sulfuric acid at 10 mA/cm². This is a preprint and laboratory result, without validation in a PEM electrolyzer.
Technology: Sequential learning, composition–property models, and LLMs
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: Lead catalysts were tested in 1 M H2SO4 at 10 mA/cm2: InMnPdOx kept overpotential below 0.5 V for over 1,000 hours; a retrospective benchmark compared the agent with Bayesian optimization and in-context language-model selection.
- Still unknown: Testing in a full PEM electrolyzer is not reported, and 'stable' means overpotential under 0.5 V rather than a broader durability assessment.
- AI role: A sequential-learning agent with ML composition models, adaptive optimization and LLM reasoning ran a human-supervised closed loop with over 90% automation.
Evidence sources: https://arxiv.org/abs/2609.30133
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
About 950 Claude agents searched a DNA-sequence database for 21 hours (210 million tokens) and flagged an uncharacterized enzyme system, found mostly in bacteriophages, with a DNA repeat array similar to CRISPR. Scientists performed the laboratory work; the function remains unknown, and the work is a preprint, not a peer-reviewed publication.
Technology: Claude (agents)
Family Claude: profile · models page
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: Human scientists expressed the protein in laboratory strains and characterized it biochemically and structurally; early experiments showed the array is expressed as distinct short RNAs.
- Still unknown: The primary function of the enzyme system is unknown; findings are preliminary, and most Claude-proposed candidates are eliminated in follow-up review.
- AI role: About 950 agents used 210 million tokens over 21 hours to search DNA sequences, narrowing about 3,500 candidate systems to 20; one agent noticed the repeat array and filed a report.
Evidence sources: https://anthropic.com/news/claude-discovers-novel-enzyme-system
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A symmetry-based search paired with molecular-dynamics calculations using a universal machine-learning force field identified five oxygen-, four lithium- and three sodium-based superionic conductor candidates. The authors report simulated diffusion properties comparable to leading materials and connected diffusion pathways. These are computational predictions; the study abstract does not establish experimental synthesis or measured ionic conductivity.
Technology: Universal machine-learning force field for molecular-dynamics screening
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: The primary abstract reports molecular-dynamics screening with a universal machine-learning force field and identifies five oxygen-, four lithium- and three sodium-based candidates with simulated diffusion comparable to leading materials.
- Still unknown: No experimental synthesis or measured conductivity is established by the reviewed abstract. The symmetry-based search idea is human-derived; the learned force field supports screening and calculated diffusion.
- AI role: A universal machine-learning force field was used in molecular-dynamics simulations that screened and evaluated the reported conductor candidates.
Evidence sources: https://pubs.acs.org/doi/10.1021/acs.chemmater.6c01341
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary publisher indexed abstract, article history, affiliation block and supporting-information listing; full article body not accessible in direct page fetch..
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A diffusion model trained on ten CMIP6 climate models estimated event-specific temperature changes conditional on observed meteorology. It found >95% probability that anthropogenic emissions since the 2015 Paris Agreement intensified Europe’s summer temperature extremes in 2021–2025; for the June 2025 heatwave, regional mean temperature is estimated to have been 0.34 °C lower under 2015 cumulative emissions, with >99% probability.
Technology: Generative diffusion model for spatially explicit temperature fields
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A diffusion model trained on ten CMIP6 climate models estimated event-specific temperature changes conditional on observed meteorology.
- Still unknown: This event-specific attribution is conditioned on observed large-scale meteorology and trained on ten CMIP6 models; it is not a direct experiment that removed emissions. The paper emphasizes uncertainty from internal variability/model spread and reports robust regional-mean effects, while grid-cell-level changes are more limited. The June 2025 value is a median modeled counterfactual estimate.
- AI role: The model learned climate-model relationships between meteorological conditions, cumulative CO₂ emissions and daily temperature fields, then generated conditional counterfactual temperature distributions to attribute the incremental post-2015 emissions effect to recent European heat extremes.
Evidence sources: https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2026GL125079
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Published in Nature, the method combines a Transformer model (R-SMILES 2) and a template-based graph model (NeuralLoc) through a learned vote. In blind tests with expert chemists, 90% of complete routes were accepted, compared with 20–50% for the component models or a baseline, and nine of ten difficult targets were solved. Evaluation covered only ten targets; one wrong prediction invalidates an entire route.
Technology: RetroChimera (Transformer and template-based model)
Family RetroChimera: profile · models page
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Blind tests by PhD-level chemists rated individual steps and complete routes for ten difficult targets: 90% of RetroChimera routes were accepted, and nine of ten targets were solved, versus 20-50% for the component models or baseline.
- Still unknown: The route-level evaluation covers only ten targets; one incorrect prediction invalidates a route; NeuralLoc is constrained on reactions outside its template library, and R-SMILES 2 may hallucinate.
- AI role: An ensemble of a Transformer (R-SMILES 2) predicting precursors and a graph neural network (NeuralLoc) selecting templates, combined by a learned rank-dependent vote.
Evidence sources: https://www.microsoft.com/en-us/research/blog/improving-synthesis-prediction-of-small-molecules-at-scale-with-retrochimera/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A literature-mining workflow using GPT-4o extracted values from 465 papers; a graph-neural-network-based pipeline screened 69,920 oxyhalide structures. Ab initio molecular-dynamics simulations for LiMoOBr₄ report a lithium content of 2.28 mol kg⁻¹, a 300 K extrapolated ionic conductivity of 13.53 mS cm⁻¹ and a 0.22 eV activation energy. The reported material properties are computational predictions, not experimental measurements.
Technology: GPT-4o literature extraction; diffusion-model and graph-neural-network screening
Family GPT-4: profile · models page
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: The paper reports computational screening of 69,920 oxyhalides and AIMD estimates for LiMoOBr₄: 2.28 mol kg⁻¹ lithium, 13.53 mS cm⁻¹ ionic conductivity extrapolated to 300 K and 0.22 eV activation energy.
- Still unknown: The checked abstract/index does not establish experimental synthesis or measured conductivity; predictions should not be described as experimentally confirmed. Full methods were not accessible in the current check.
- AI role: GPT-4o was used to extract literature data from 465 papers; a learned graph-neural-network screening pipeline contributed to nomination of LiMoOBr₄ for computational validation.
Evidence sources: https://www.sciencedirect.com/science/article/pii/S3051383926001210
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary publisher indexed abstract/highlights and saved publisher affiliation metadata; direct publisher page returned 403. University announcement checked only as corroboration, not as primary evidence..
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The authors prove that, assuming Zieve-admissible branch data are realizable for three branch points, all such data with four or more branch points are realizable. This establishes that no additional obstruction arises solely from adding branch points; the underlying three-branch-point conjecture remains open. The paper discloses GPT-5.6 Sol and GPT-6 Astra assistance in exploring reduction strategies and in proof development and checking.
Technology: OpenAI GPT-5.6 Sol and GPT-6 Astra
Family GPT-5: profile · models page
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: The authors prove that, assuming Zieve-admissible branch data are realizable for three branch points, all such data with four or more branch points are realizable.
- Still unknown: This is a new conditional mathematical reduction theorem, not a resolution of Zieve’s full conjecture: the three-branch-point case remains open. The source is an arXiv preprint; human authors report reviewing AI-assisted work and accept responsibility, but no independent verification of the proof is claimed here.
- AI role: Authors explicitly report that the models assisted throughout the research, including exploration of reduction strategies and proof development/checking, as well as notation consistency and manuscript work. Humans reviewed all AI-assisted content and take responsibility for the proof.
Evidence sources: https://arxiv.org/abs/2609.23663
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Using 6.3 fb⁻¹ of 13.6 TeV proton-proton collisions recorded in 2024, LHCb observed an Ωc⁰π⁺ mass peak with 8.7σ global significance and identified it as the weakly decaying doubly charmed Ωcc⁺ baryon. The reported mass is 3725.9 ± 1.0 (stat) ± 0.2 (syst) ± 0.4 (lifetime) ± 0.6 (external) MeV/c². A boosted decision tree suppressed random-track combinatorial background; the result is a preprint submitted to Physical Review Letters.
Technology: Boosted decision tree classifier (TMVA toolkit)
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: Using 6.3 fb⁻¹ of 13.6 TeV proton-proton collisions recorded in 2024, LHCb observed an Ωc⁰π⁺ mass peak with 8.7σ global significance and identified it as the weakly decaying doubly charmed Ωcc⁺ baryon.
- Still unknown: The source is an LHCb collaboration preprint submitted to PRL, not yet a peer-reviewed journal publication as of its arXiv version. The observed peak has 8.7σ global significance; the mass uncertainty includes an explicit lifetime-dependent selection-bias term because the Ωcc⁺ lifetime is unknown. BDT use is an analysis component, not the sole basis of the observation.
- AI role: The BDT was trained on simulated Ωcc⁺ signal decays and wrong-charge Ωc⁰π⁻ combinations from data and used to suppress the dominant random-track combinatorial background in the offline candidate selection. The background rejection supported the mass-spectrum analysis yielding the first observation.
Evidence sources: https://arxiv.org/abs/2609.21921
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
In just under four weeks, Claude optimized over 30 models for structure, protein design, and genomics. They were on average four times faster with a small accuracy loss, and nearly twice as fast with identical outputs. It also created a low-memory mode for systems over 10,000 tokens on a single GPU node. Design results are in silico only, and comparisons are internal to Anthropic.
Technology: Claude (in Claude Science)
Family Claude: profile · models page
Publication: Announcement only · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: retrospective or benchmark
- Tested: Accelerated models were checked against their downstream tasks (for example structure prediction) and FlashPairformer kernels were compared with cuEquivariance (2.7-2.9x faster on triangle attention); design results are in silico with internal comparisons.
- Still unknown: Design results have no wet-lab validation, large viral capsid predictions were incorrect, ColabFold kernels were not benchmarked, and the model is not reproducible.
- AI role: Claude wrote custom FlashPairformer kernels, optimized more than 30 models, built a low-memory Big mode, and ran autonomous protein binder design.
Evidence sources: https://www.anthropic.com/research/claude-uplifts-biomolecular-modeling
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The paper gives a negative answer, already for q=2, to Talagrand’s question whether an operator’s Rademacher cotype is universally bounded by its Gaussian cotype and its (q,1)-summing norm. For every sufficiently large power-of-two dimension it constructs a Banach space and operator violating such a universal bound. The same construction also yields a negative answer to Kwapień’s decomposition problem.
Technology: ChatGPT GPT-5.6
Family GPT-5: profile · models page
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: The paper gives a negative answer, already for q=2, to Talagrand’s question whether an operator’s Rademacher cotype is universally bounded by its Gaussian cotype and its (q,1)-summing norm.
- Still unknown: This is a single-author arXiv preprint, not a peer-reviewed publication. GPT-5.6 is credited with discovering the counterexample during discussions; the author says they independently verified all arguments. No Lean or other proof-assistant formalization is reported. The mathematical scope is the stated q=2 counterexample and its stated corollary, not a complete characterization of operator cotype.
- AI role: The author states that the counterexample was discovered during discussions with GPT-5.6. The paper presents an explicit Walsh–Hadamard-based construction and proof; the author independently verified the arguments.
Evidence sources: https://arxiv.org/abs/2609.19731
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Paper2Agent-generated agents used AlphaGenome to prioritize GPR137 as the top predicted gene affected by psoriasis-associated variant rs887314 in CD4+ T cells, then integrated that prediction with existing MPRA-coupled scCRISPRi and CD4+ T-cell Perturb-seq data. A cross-screen correlation of gene-expression signatures supported GPR137 as the probable causal gene under stimulated conditions (Spearman rho 0.613 at 8 h and 0.630 at 48 h; FDR < 0.05), but not at rest. The paper presents this as activation-dependent computational evidence.
Technology: Paper2Agent-generated AlphaGenome, MPRA-coupled scCRISPRi and CD4+ T-cell Perturb-seq agents, coordinated through an AI co-scientist
Family AI Co-Scientist: profile · models page
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Paper2Agent-generated agents used AlphaGenome to prioritize GPR137 as the top predicted gene affected by psoriasis-associated variant rs887314 in CD4+ T cells, then integrated that prediction with existing MPRA-coupled scCRISPRi and CD4+ T-cell Perturb-seq data.
- Still unknown: The analysis reuses existing published datasets and does not report new wet-lab experiments for this psoriasis case. AlphaGenome supplied a model prediction; cross-dataset signature correlation supports, but does not establish, causal biology. The GPR137 concordance was significant only after stimulation, and the authors describe it as probable/activation-dependent. A human researcher selected which AI-proposed strategy to run. Primary paper source checked; no independent replication assessed.
- AI role: The AlphaGenome paper agent scored and ranked candidate genes at rs887314. Under human-in-the-loop supervision, the AI co-scientist inspected two other paper agents' datasets, proposed ten validation strategies, and the researcher selected signature-correlation analysis. The resulting cross-screen, cross-modality analysis found significant concordance for GPR137 only under stimulated conditions.
Evidence sources: https://www.nature.com/articles/s41586-026-11044-y
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A counting argument proves that every finite graph with average degree greater than t−2 contains every tree on t vertices, resolving the Erdős–Sós conjecture. The proof counts adjacency-marked vertex orderings and uses an induction on the tree to bound those not supporting a rooted copy. The exposition states the sharp classical theorem and records its proof was discovered by GPT-6 Astra; a Lean repository checks the benchmark formalization of the result. Related expository papers of the same proof are aliases, not separate discoveries.
Technology: Pre-release OpenAI GPT-6 Astra in the FrontierMath Erdős benchmark
Family GPT-6: profile · models page
Publication: Preprint · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: prospective or independent check
- Tested: A counting argument proves that every finite graph with average degree greater than t−2 contains every tree on t vertices, resolving the Erdős–Sós conjecture.
- Still unknown: The arXiv paper is an exposition of the proof, not the original benchmark output. The linked Lean proof is machine-checked against the benchmark formalization; the registry notes a minor parity-case difference in that compared statement, while the paper’s counting lemma derives the sharp classical average-degree threshold. This remains a preprint and the formalization is not independently audited here.
- AI role: GPT-6 Astra autonomously found the counting proof while solving Erdős problem 548 in the FrontierMath Erdős benchmark; the exposition author reports the discovery and lays out the proof. A Lean 4 repository states the proof was machine-checked by a Comparator against the benchmark’s formal statement.
Evidence sources: https://arxiv.org/abs/2609.17877
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 4/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A graph variational autoencoder proposes molecules while learned classifiers and regressors, retrosynthesis and stock-inventory constraints guide selection. The journal version reports 11 first-generation materials characterized through mixture-based extrapolation and two additional materials selected in a second iteration that exhibited the ferroelectric-nematic phase in their neat state. The first arXiv version (2025-12-18) contained the first-generation mixture-based work, not the later neat-state confirmation; this draft uses the journal date for that result. Predictions have material errors, and the work does not establish device performance.
Technology: Graph variational autoencoder; graph neural-network ensembles; XGBoost
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Journal full text reports two second-round synthesized materials with neat-state ferroelectric-nematic phase; first 11 compounds were assessed through mixture-based extrapolation.
- Still unknown: Transition predictions may differ substantially from observed values; independent replication and device performance not established.
- AI role: Generative and predictive graph models rank and propose molecules; inventory and retrosynthetic filters guide laboratory synthesis and characterization.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13569083/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 16 Nov 2026. Review scope: Primary arXiv v1 PDF directly read and primary journal full text indexed on PMC; version-specific first-cohort versus second-round evidence reconciled..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Combining marine natural-product fragment mining, AI-based molecular generation and structure-guided selection, the authors designed optimized analogues Ib18 and It12 targeting the ABC transporters P-glycoprotein (ABCB1) and BCRP (ABCG2). Both analogues reversed transporter-mediated resistance in overexpressing cells and outperformed reference inhibitors. Target-engagement assays supported transporter inhibition without expression downregulation. In xenografts, Ib18 increased mitoxantrone antitumor activity and tumor accumulation without a detected increase in systemic toxicity.
Technology: AI-based molecular generation integrated with marine natural-product fragment mining and structure-guided prioritization
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Target-engagement assays supported transporter inhibition without expression downregulation.
- Still unknown: The evidence is cellular and preclinical xenograft evidence, not patient efficacy or established clinical safety. The abstract reports no systemic toxicity increase in the studied xenograft setting, which should not be generalized. It does not give quantitative potency values or detailed comparator identities in the abstract.
- AI role: AI-based generation contributed candidates to a natural-product-inspired library; structure-guided prioritization selected optimized analogues for synthesis and functional tests. The authors tested Ib18 and It12 in transporter-overexpressing cells and evaluated Ib18 with mitoxantrone in xenografts.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/42720471/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A robotic active-learning campaign using SARA-H mapped processing regions that stabilize δ-Bi₂O₃ and Bi₂Ti₂O₇ in thin films and examined dwell-dependent phase behavior in the Bi–Ti–O system. The study also experimentally supported the prediction that bismuth substitution in TiO₂ inhibits the anatase-to-rutile transformation. This is a human-in-the-loop experimental materials result; the compounds/phases were not all newly discovered, while the mapped processing domains and comparative evidence are the study-specific contribution.
Technology: SARA-H (Scientific Autonomous Reasoning Agent with human input) and automated probabilistic phase labeling
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: A robotic active-learning campaign using SARA-H mapped processing regions that stabilize δ-Bi₂O₃ and Bi₂Ti₂O₇ in thin films and examined dwell-dependent phase behavior in the Bi–Ti–O system.
- Still unknown: This is a published, human-guided experimental study, but the phases themselves are not claimed as newly discovered compounds. The study-specific finding is experimentally mapped processing/dwell behavior in the Bi–Ti–O campaign and evidence that Bi substitution inhibits anatase-to-rutile transformation. Scope is thin films and tested processing conditions, not bulk phase stability or universal behavior.
- AI role: Used phase-identification outputs to select and interpret robotic laser-annealing experiments, with human domain guidance improving search efficiency.
Evidence sources: https://journals.aps.org/prxintelligence/abstract/10.1103/c362-349b
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The paper proves that for every finite family of Euclidean-norm-at-most-one vectors, signs can be chosen so the signed sum has infinity norm below 3√(2π), independent of vector count and dimension. It derives a corresponding square-root-in-sets bound for Beck–Fiala discrepancy. The authors explicitly credit the Odin Automatic AI Research Agent with discovering the proof; a separate elementary proof was posted independently. The algorithmic construction of such signings is a separate follow-up result and is not claimed here.
Technology: Odin Automatic AI Research Agent
Publication: Preprint · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: prospective or independent check
- Tested: The paper proves that for every finite family of Euclidean-norm-at-most-one vectors, signs can be chosen so the signed sum has infinity norm below 3√(2π), independent of vector count and dimension.
- Still unknown: The result is reported in an arXiv preprint, not a peer-reviewed journal source in this review. A separate elementary proof of the conjecture appeared later, providing an independent proof route; the paper’s AI attribution is author-reported. The result establishes existence, not a polynomial-time signing algorithm.
- AI role: The authors credit Odin with discovering the proof of their vector-balancing theorem; Duke’s official account says the system discovered the proof and human authors scrutinized and organized the argument into the paper.
Evidence sources: https://arxiv.org/abs/2609.11189
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 4/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A precomputed database of predicted molecular effects for about 9 billion possible single-letter changes in the human genome, generated by AlphaGenome. Researchers can access it free through a portal and API. The predictions are a research tool, do not replace experiments, and are not validated for clinical use.
Technology: AlphaGenome, AlphaGenome Variant Impact (AVI)
Family AlphaGenome: profile · models page
Publication: Announcement only · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: retrospective or benchmark
- Tested: Validation described is the company's own testing of the AlphaGenome Variant Impact score on pathogenicity benchmarks, plus case studies: a DNM1 variant prioritized with GREGoR and 22% more non-coding associations in UK Biobank.
- Still unknown: The page does not detail how individual predictions were checked; AlphaGenome is not validated or approved for clinical use.
- AI role: AlphaGenome predicted variant effects across cell types and tissues; these were precomputed for about 9 billion variants and combined with AlphaMissense into a single score.
Evidence sources: https://deepmind.google/blog/alphagenome-atlas-a-predictive-map-of-every-possible-dna-letter-change-in-the-human-genome/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
OpenAI announces a proof that three-dimensional solutions to the Navier–Stokes equations can develop a singularity in finite time and publishes a text formalized in Lean. This is the source's claim: this record does not establish community consensus, correctness of the artifacts, or recognition as a Millennium Prize problem.
Technology: OpenAI internal model
Publication: Announcement only · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: prospective or independent check
- Tested: OpenAI shared an analytical proof and a Lean formalization that an initially smooth fluid at rest can develop a finite-time singularity under a smooth force with finite energy.
- Still unknown: Not stated in the reviewed material
- AI role: An internal OpenAI model, described as more capable than GPT-6 Astra, produced the proof and the Lean formalization.
Evidence sources: https://openai.com/index/navier-stokes-solution/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 5/5 · Validation 4/5 · Novelty 5/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Connected to MIT EQuS lab software, GPT-5.6 Sol selected parameters, ran measurements, and calibrated a six-qubit superconducting chip, measuring its frequencies and coherence. The demonstration covers known procedures on real hardware; weak or noisy signals sometimes required expert assistance, and no new physical property was demonstrated.
Technology: GPT-5.6 Sol and Codex
Family GPT-5: profile · models page
Publication: Announcement only · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: direct test
- Tested: Connected to MIT EQuS lab software, GPT-5.6 Sol ran measurements on an uncalibrated six-qubit chip of a standard type, chose parameters, operated the hardware and analyzed results.
- Still unknown: Not assessed: the technical case study with the detailed results was not read.
- AI role: GPT-5.6 Sol in Codex ran routine measurement workflows with measurement-specific skills provided by the researcher.
Evidence sources: https://openai.com/index/codex-quantum-computing-experiments/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The authors prove the classical complex Grothendieck constant exceeds 1.35584631827168, closing more than one quarter of the gap between Davie’s lower bound and Haagerup’s upper bound reported in the paper. The numerical search used Hermite projections and a radial weight; interval arithmetic on exact rational inputs rigorously verifies the stated bound. The authors credit the Odin Automatic AI Research Agent with discovering both the lower bound and its proof.
Technology: Odin Automatic AI Research Agent
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: The authors prove the classical complex Grothendieck constant exceeds 1.35584631827168, closing more than one quarter of the gap between Davie’s lower bound and Haagerup’s upper bound reported in the paper.
- Still unknown: The source is an arXiv preprint. Its stated lower bound is computer-assisted and the proof includes interval arithmetic with exact inputs; the paper does not determine the exact complex Grothendieck constant. The authors’ AI attribution is not independent evidence about the system’s internal process.
- AI role: The authors explicitly state that Odin discovered the improved lower bound and its proof; the preprint describes numerical optimization that found a candidate, followed by exact rational inputs and interval-arithmetic verification by the authors’ supplied code.
Evidence sources: https://arxiv.org/abs/2609.07000
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A proof of Fermat's Last Theorem fully machine-verified in Lean: Claude produced most of it independently over 11 days, with about 13 million lines and 29,500 intermediate theorems, and a comparator confirms that its statement matches Mathlib's. It follows a simplified version of Wiles's proof, uses Mathlib and parts of an Imperial College project, and is probably longer than necessary.
Technology: Claude, Lean
Family Claude: profile · models page
Publication: Announcement only · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: prospective or independent check
- Tested: The proof was checked by Lean using only standard axioms; a comparator tool confirmed the theorem statement matches Mathlib's, and mathematician Kevin Buzzard reviewed it.
- Still unknown: The proof is about 13 million lines and likely much longer than needed; it follows a simplified version of Wiles's proof and is presented as verification, not new mathematics.
- AI role: Dozens of agents worked for 11 days on the Prove2Me platform, producing about 30,300 computer-verified theorems; humans gave occasional high-level instructions.
Evidence sources: https://www.anthropic.com/research/formalizing-fermats-last-theorem
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 4/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A YOLO object-detection model trained on 2023 camera-trap videos detected European hamsters with an F1-score of 0.93 and accuracy of 0.93. Its activity profile matched human annotation and showed peak activity at dusk and a secondary peak before sunrise. The evidence covers one season at one monitoring setup, so generalization to other sites is untested.
Technology: YOLO object detection (deep learning)
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A YOLO detector trained on 2023 camera-trap videos was tested for European hamster detection (F1 and accuracy 0.93), and its activity profile was compared with human annotation.
- Still unknown: The abstract states no limitations; it covers one season and does not name the sites or the number of cameras.
- AI role: A YOLO object-detection model detected hamsters in camera-trap footage.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1002/ece3.74267&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A model that learns from real-time geostationary satellite data and produces hourly forecasts, at resolutions down to 5 km for temperature and humidity (about five times finer than WeatherNext 2). For short-term precipitation, Google reports improvements of up to 60% over NASA's IMERG. These are ‘up to’ figures from Google's evaluations, with an independent ranking by Brightband.
Technology: WeatherNext 3
Family WeatherNext: profile · models page
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: Precipitation forecasts were scored with CRPS against baselines: gains of up to 60% over IMERG, 30% over MRMS and 10% over rain gauges for early lead times; independent live evaluations by Brightband ranked the model.
- Still unknown: The 'up to' figures lack full test setup, longer-range gains are largest where forecasts are weak, and the atmosphere retains unpredictability.
- AI role: The model ingests live geostationary satellite data and historical analyses, trains on station observations and produces gridded fields, cyclone tracks and station-level forecasts.
Evidence sources: https://blog.google/innovation-and-ai/models-and-research/google-deepmind/introducing-weathernext-3/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Using an expanded Pn-wave travel-time database, the authors construct a high-resolution uppermost-mantle velocity and anisotropy model for the southwestern United States. The tomographic interpretation identifies separate low-velocity, anisotropic upwelling conduits beneath the Black Rock Desert, Uinkaret, and San Francisco volcanic fields along the Colorado Plateau margins, and supports remnants of the ancient Farallon slab beneath the central Great Basin.
Technology: Seis-PickNet deep-learning seismic phase picker
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Using an expanded Pn-wave travel-time database, the authors construct a high-resolution uppermost-mantle velocity and anisotropy model for the southwestern United States.
- Still unknown: These are tomographic interpretations of seismic velocity and anisotropy, not direct images or direct measurements of magma conduits. The article abstract reports inferred independent thermal upwellings and supports (rather than proves uniquely) Farallon slab remnants. Resolution and tectonic interpretation depend on the travel-time picks and inversion assumptions; the paper's abstract does not establish direct volcanic plumbing connectivity.
- AI role: The authors used Seis-PickNet to automatically pick Pn phases and expand the regional travel-time database; those new picks underpin the velocity and anisotropy tomography from which the mantle structures are interpreted. The AI supplies the expanded seismic measurements used in the scientific inversion, rather than only classifying existing results.
Evidence sources: https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/2026JB034108
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
ERTransformer generated epitope-conditioned TCR β-chain CDR3 candidates. Flow-cytometry experiments found selected artificial TCRs induced activation comparable to or greater than natural controls in defined TCR/pMHC contexts. This represents the September 2026 journal report; an ERTransformer precursor was posted October 20, 2023. The earlier report does not establish that these journal wet-lab results were already public.
Technology: ERTransformer; EpitopeBERT; ReceptorBERT
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Selected generated candidates tested for T-cell activation by flow cytometry.
- Still unknown: Full experimental methods and supplements not assessed: publisher rendering exposed abstract and metadata. General specificity, safety and clinical outcomes are not established by this review.
- AI role: Generates TCR β-chain CDR3 candidates conditioned on epitopes.
Evidence sources: https://link.springer.com/article/10.1186/s13059-026-04266-w · https://www.biorxiv.org/content/10.1101/2023.10.18.562845v1
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
MAPL-EMIT used a vision transformer trained on 3.6 million synthetic plumes embedded in EMIT radiance data to map methane sources. Beyond synthetic tests, the paper abstract reports recovery of 84% of known annotated plume complexes across 1,084 EMIT granules. Comparisons with airborne data, landfills and controlled releases supported previously uncaptured sources. This is source-reported validation, not an independent emissions audit.
Technology: MAPL-EMIT (end-to-end vision transformer on EMIT hyperspectral radiance)
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Synthetic tests plus known real plumes, coincident airborne observations, landfill comparisons and controlled-release data.
- Still unknown: Review scope: complete indexed abstract and metadata; full uncertainty budgets and individual source emissions were not assessed.
- AI role: An end-to-end vision transformer jointly estimated methane enhancements, plume outlines and source locations from hyperspectral radiance.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.1073%2Fpnas.2612145123%22&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 2/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Profiling 5,416 plasma proteins in Korean cohorts, the study found distinct signatures across four biopsy-proven primary glomerular diseases. An elastic-net model using 93 proteins differentiated disease subtypes and controls in discovery and external validation cohorts. It achieved AUCs above 0.8 for minimal change disease, membranous nephropathy and IgA nephropathy; the model correctly classified 14 of 15 minimal-change cases and 20 of 32 IgA-nephropathy cases, but only 3 of 14 focal-segmental glomerulosclerosis cases. Functional analysis highlighted hemostasis-related proteins in minimal change disease.
Technology: Elastic-net-regularized logistic regression using plasma proteomic features
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Profiling 5,416 plasma proteins in Korean cohorts, the study found distinct signatures across four biopsy-proven primary glomerular diseases.
- Still unknown: The discovery cohort contained 147 participants and the external validation cohort 85; subtype-level validation samples were small. Classification was strong for minimal change disease, membranous nephropathy and IgA nephropathy but poor for focal segmental glomerulosclerosis (3/14 correctly classified). The findings support cohort-level proteomic differentiation and do not establish a clinically deployable noninvasive replacement for kidney biopsy or causal roles for the selected proteins.
- AI role: The model used elastic-net regularization to select and combine a 93-protein plasma signature for classifying biopsy-proven primary glomerular disease subtypes and controls. The model was evaluated in an independent external validation cohort. Functional analysis of key proteins was used to characterize subtype-associated pathways, including hemostasis in minimal change disease.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/41779447/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
ESM-FEP was applied to Zea mays dioxygenase ZmHSL1B and nominated a quadruple mutant (Q140H/Y205F/L332R/K336F) with approximately sevenfold higher catalytic efficiency for mesotrione detoxification than wild type, measured by in-vitro assays and kinetic analysis. Arabidopsis thaliana expressing the engineered variant showed enhanced herbicide tolerance in the reported transgenic tests.
Technology: ESM-FEP (fine-tuned ESM-2 protein language model plus alchemical free-energy perturbation)
Family ESM / ESMFold: profile · models page
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: ESM-FEP was applied to Zea mays dioxygenase ZmHSL1B and nominated a quadruple mutant (Q140H/Y205F/L332R/K336F) with approximately sevenfold higher catalytic efficiency for mesotrione detoxification than wild type, measured by in-vitro assays and kinetic analysis.
- Still unknown: Reported results concern one maize dioxygenase and mesotrione in specified in-vitro and Arabidopsis tests. They do not establish crop-field performance, commercial efficacy, or generality across herbicides/enzymes. Primary PubMed abstract and publisher citation metadata checked; no independent replication assessed.
- AI role: Fine-tuned ESM-2 screened saturation-mutagenesis variants for retained global protein fitness; free-energy perturbation then estimated substrate-binding effects and prioritized mutants. The selected M5 mutant was assayed in vitro and in transgenic Arabidopsis by the researchers.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/42675858/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A single-author preprint proves that every finite one-dimensional open O(n) spin chain with arbitrary inhomogeneous nearest- and next-nearest-neighbor interactions has an exact, temperature-independent Hamiltonian mapping to a simpler nearest-neighbor chain with axial single-spin potentials. A parallel result for q-state Potts chains yields a closed-form solution for every q ≥ 2 and finite chain length. The theorem generalizes a 1990 Heisenberg-chain challenge and connects spontaneous collective behavior from competing interactions to field-induced behavior.
Technology: OpenAI GPT-5.6 Sol; GPT-5.6 Sol Pro for final adversarial manuscript review
Family GPT-5: profile · models page
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: A single-author preprint proves that every finite one-dimensional open O(n) spin chain with arbitrary inhomogeneous nearest- and next-nearest-neighbor interactions has an exact, temperature-independent Hamiltonian mapping to a simpler nearest-neighbor chain with axial single-spin potentials.
- Still unknown: This is a single-author arXiv preprint, not yet peer reviewed. The theorem concerns classical one-dimensional open chains and exact Hamiltonian mappings; the paper discusses possible experimental relevance but reports no experimental test. The author assumes responsibility for reviewing and verifying AI-generated content; Mathematica 14.3 was also used to verify results.
- AI role: The author credits GPT-5.6 Sol with scientific discovery, including generating the recursive Householder moving-frame proof and recognizing broad generalizations. The author identified the hidden reciprocity, extended the theorem to arbitrary spatial inhomogeneity, and states that all AI-generated scientific content was reviewed and verified. GPT-5.6 Sol Pro conducted a final adversarial review.
Evidence sources: https://arxiv.org/abs/2609.00322
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Researchers built PIRAG-LM, which combines a structured knowledge base of experimental synthesis routes, physics-based chemical/structural/thermodynamic similarity retrieval, and LLM route reasoning. In prospective tests, it guided synthesis of five compounds previously unreported in experimental literature: BaMo0.3In0.7O2.95, BaNb0.4In0.6O2.9, Hg[B(CN)4]2, CoCo(CN)6 and SrNb2Fe2(PO4)6. Powder X-ray diffraction patterns were consistent with target phases. This is one paper-level synthesis-planning result, not five separate entries.
Technology: PIRAG-LM, physics-informed retrieval-augmented language model, with DeepSeek-V4 used in the structured synthesis-knowledge-base extraction workflow
Family DeepSeek: profile · models page
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: Researchers built PIRAG-LM, which combines a structured knowledge base of experimental synthesis routes, physics-based chemical/structural/thermodynamic similarity retrieval, and LLM route reasoning.
- Still unknown: This is a source-checked preprint (not treated as peer-reviewed here). The reported new-compound claims refer to absence from experimental literature and prospective synthesis; PXRD matches support target phase formation, but the paper does not establish broad application performance or independent replication. Experts selected materials and adjusted operational conditions; the system did not autonomously run experiments. Five compounds remain bundled as one paper-level finding.
- AI role: PIRAG-LM retrieved synthesis precedents by chemical, structural and thermodynamic similarity and proposed synthesis methods, precursors and processing conditions. Experts evaluated routes and refined practical parameters; researchers synthesized and characterized the five target phases. The final route class, precursor logic and main processing sequence generally followed model recommendations.
Evidence sources: https://arxiv.org/abs/2608.25392
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A knowledge-informed graph attention model with uncertainty quantification selected the KG-AMS nickel superalloy composition and processing conditions for laser directed-energy deposition. The study reports experimental validation with defect area fractions of 0.037%–0.133% and hardness of 391–417 HV across tested parameter sets; it also reports room-temperature tensile strength of 1307 MPa and elongation of 22.2%. These measurements apply to the tested alloy and processing window, and larger builds remain untested. This entry covers KG-AMS only; the paper’s separate aluminum alloy KG-AMAA had already been reported by the same group in 2025.
Technology: Knowledge-informed graph attention network with uncertainty quantification (UQ-KGAT)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The paper reports fabricated KG-AMS nickel alloy specimens with defect area fractions of 0.037%–0.133% and hardness of 391–417 HV over the reported processing parameters; it also reports room-temperature tensile strength 1307 MPa and elongation 22.2%.
- Still unknown: Results are bounded to the specimens and processing conditions in the study; larger builds and independent replication are not established. The separate KG-AMAA aluminum alloy is not counted here because the same group reported it earlier.
- AI role: UQ-KGAT, a knowledge-informed graph attention network with uncertainty quantification, identified the KG-AMS alloy and its processing region for experimental validation.
Evidence sources: https://www.nature.com/articles/s41467-026-77119-6 · https://pmc.ncbi.nlm.nih.gov/articles/PMC13612592/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary Nature Communications article, linked supplementary affiliation block, and indexed PMC full text checked for date, alloy/model link and tested properties; record is limited to KG-AMS to avoid duplication of KG-AMAA..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Researchers trained convolutional neural networks on E11.5 mouse heart, limb and midbrain chromatin-accessibility data, then fine-tuned the models using validated human and mouse enhancers and used model-guided sequence optimization to design 15 synthetic enhancers. In vivo transgenic reporter assays found all 15 reproducibly active in their intended tissue. Four of five heart designs and four of five CNS designs were exclusive to the target tissue; all five limb designs were strongly active in limb and had weaker activity in other TWIST1-expressing mesenchymal tissues.
Technology: Convolutional neural networks for sequence-to-accessibility and sequence-to-activity prediction, transfer learning, and Ledidi gradient-based sequence design
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Researchers trained convolutional neural networks on E11.5 mouse heart, limb and midbrain chromatin-accessibility data, then fine-tuned the models using validated human and mouse enhancers and used model-guided sequence optimization to design 15 synthetic enhancers.
- Still unknown: The functional test was a transgenic reporter assay at embryonic day 11.5, not editing or activation of an endogenous native gene locus. Fifteen selected sequences were tested (five per target), so the reported all-positive result does not estimate success across every generated candidate. Some designs had weaker off-target activity, particularly in related mesenchymal tissues. These results do not establish therapeutic efficacy or safe gene therapy.
- AI role: Models learned tissue-specific chromatin-accessibility patterns from genome-wide ATAC-seq, then transfer learning adapted them to validated enhancer activity. The trained models guided de novo enhancer sequence edits; the authors selected 15 candidates for transgenic mouse-embryo reporter testing, where each showed reproducible activity in its intended target tissue.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13553301/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Researchers used deep learning with iterative manual optimization to identify 174,929 PKP precursors in more than two million earthquake waveforms recorded from 1990 to 2024. The expanded global map connected previously isolated sampling regions and highlighted six high-potential scattering zones near the core–mantle boundary that the authors report had not previously been documented. These are model-inferred targets for later seismic imaging, not fully characterized structures.
Technology: Deep-learning PKP precursor detector with human iterative optimization
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Researchers used deep learning with iterative manual optimization to identify 174,929 PKP precursors in more than two million earthquake waveforms recorded from 1990 to 2024.
- Still unknown: The six features are high-potential scattering zones inferred from probability maps and proposed as targets for future multi-phase inversions; they are not precise outlines or independently characterized structures. The proposed thermochemical explanations are interpretations. Cross-validation with independent seismic phases supports multiple mapped ULVZs, but does not individually validate all six newly highlighted zones.
- AI role: The deep-learning detector screened over two million earthquake waveforms to identify 174,929 weak PKP precursor signals; researchers iteratively optimized detections manually. This AI-enabled catalog increased global sampling density by an order of magnitude and supported the dual-probability map that identified six previously undocumented high-potential scattering zones.
Evidence sources: https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2025JB033195
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A closed-loop Bayesian experimental-design method, BEAM, found multiple defect-free directed-energy-deposition configurations for GRCop-42 copper alloy at infrared laser powers from 950 down to 500 W. Researchers report that these experimentally successful settings solved a process-development problem that had resisted several months of manual trials, enabling GRCop-42 fabrication on lower-power platforms. This is a manufacturing-process result for an existing alloy, not a new alloy composition.
Technology: BEAM (Bayesian Experimental design for Additive Manufacturing)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: A closed-loop Bayesian experimental-design method, BEAM, found multiple defect-free directed-energy-deposition configurations for GRCop-42 copper alloy at infrared laser powers from 950 down to 500 W.
- Still unknown: The paper establishes feasible configurations in one directed-energy-deposition system for one alloy. It does not demonstrate aerospace-qualified components, service life or broader transfer to other machines and alloys. The reported defect-free characterization is bounded to the study's print assessment.
- AI role: BEAM uses a probabilistic surrogate and adaptive experimental-design strategy to select small batches of process configurations for laboratory testing; print outcomes update its feasibility model and guide subsequent tests. The loop identified successful configurations at 950, 700, 600 and 500 W, where prior manual trials had failed for months.
Evidence sources: https://onlinelibrary.wiley.com/doi/10.1002/aaai.70069
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
MACE-MP-0b2 interatomic-potential calculations screened 462 four- and five-cation trivalent-oxide compositions and selected 16 for synthesis. Experiments produced three phase-pure corundum high-entropy oxides—(Al,Cr,Fe,Ga)₂O₃, (Al,Cr,Fe,Rh,Sc)₂O₃ and (Cr,Fe,Ga,Rh,Sc)₂O₃—and a separate phase-pure cation-ordered (Al,Fe,Ga,Sc)₂O₃ oxide. The ordered compound crystallized in orthorhombic Pba2; its precise cation-site occupancies were not resolved conclusively by powder diffraction.
Technology: MACE-MP-0b2 machine-learning interatomic potential (MLIP)
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: MACE-MP-0b2 interatomic-potential calculations screened 462 four- and five-cation trivalent-oxide compositions and selected 16 for synthesis.
- Still unknown: This is an unrefereed arXiv preprint. Only three of 16 tested compositions formed phase-pure corundum high-entropy oxides; most attempts produced competing phases, and rhodium-containing samples showed redox instability. Synthesis outcomes depended strongly on route: the three corundum HEOs were phase-pure only with one of solid-state or combustion synthesis. The ordered oxide’s structure is supported by powder diffraction/Rietveld refinement, but exact site preferences could not be conclusively determined from pXRD. The screening supports candidate ranking, not reliable prediction of phase purity or synthesis conditions for arbitrary HEOs.
- AI role: The MLIP relaxed approximately 1,000-atom disordered supercells and estimated mixing enthalpy and atom-energy-distribution descriptors for the 462-composition search. The authors combined these with a stability metric to rank 16 synthesis candidates. Experimental powder-X-ray diffraction and Rietveld refinement then assessed the predicted materials, linking model-guided selection to the new synthesized compositions.
Evidence sources: https://arxiv.org/abs/2608.20596
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
After an initial prompt, Claude selected binding sites, used design and folding models, and chose candidates, with success on 14 of 15 targets. Of 1,320 designs, 354 bound, for rates of 22–35% versus a typical 10–15%. None of the 90 designs for the maltose-binding protein worked. External labs ran the tests.
Technology: Claude (Mythos Preview and Opus 4.8)
Family Claude: profile · models page
Publication: Announcement only · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: direct test
- Tested: External labs (Adaptyv Bio and Twist Bioscience) produced and tested the designs: 354 of 1,320 designs bound, with hit rates of 22-35% versus a typical 10-15%; none of 90 designs for maltose-binding protein bound. The collaborating external laboratories performed candidate production and binding assays within the discovery project; separate independent follow-up replication is not established by this account.
- Still unknown: Anthropic plans further characterization to confirm hit rates and affinities; BBF-14 affinities are modest, GDF-8 data were inconclusive and excluded, and the TNFa result is unexplained.
- AI role: Claude chose binding sites, generated candidate structures and sequences by orchestrating structure, sequence and co-folding models, ran in silico optimization and chose candidates.
Evidence sources: https://www.anthropic.com/research/Claude-accelerates-protein-design
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A closed-loop materials workflow combined an LLM that distilled synthesis knowledge from papers, high-throughput hyperspectral imaging, and multi-objective Bayesian optimization with experimental feedback. In a demonstration, the team synthesized previously unreported Rb3BiI6 thin films and validated optimized films with optical bandgap analysis and X-ray diffraction. LLM-assisted initialization produced more Pareto-optimal samples and higher hypervolume than matched Latin-hypercube initialization in the reported paired campaign.
Technology: LLM-assisted synthesis-knowledge extraction and initialization with multi-objective Bayesian optimization and high-throughput hyperspectral imaging
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: A closed-loop materials workflow combined an LLM that distilled synthesis knowledge from papers, high-throughput hyperspectral imaging, and multi-objective Bayesian optimization with experimental feedback.
- Still unknown: This is a preprint. The specific Rb3BiI6 result is reported as a previously unreported compound experimentally synthesized as thin films and checked by XRD and optical bandgap analysis; the study does not establish independent replication, long-term stability or device performance. Do not treat the accompanying optimization benchmark as an independent discovery.
- AI role: The LLM distilled literature synthesis expertise and initialized candidate experimental choices; Bayesian optimization iteratively selected subsequent conditions based on measured hyperspectral outcomes. Human operators synthesized thin films and performed optical/XRD validation.
Evidence sources: https://arxiv.org/abs/2608.15928
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
FluEmbed, a protein language model framework, predicted influenza A (H3N2) antigenicity from RNA sequence with Spearman correlations of 0.67 to 0.80 against hemagglutination inhibition titers. It outperformed Hamming, BLOSUM62 and phylogenetic baselines, and in-silico mutagenesis showed unconstrained mutations carry greater escape potential outside dominant lineages. Validation uses existing HI assay data and retrospective sequences; no new wet-lab tests were run.
Technology: FluEmbed (protein language model embeddings)
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: FluEmbed predictions from H3N2 sequences were correlated with existing hemagglutination inhibition titers (Spearman 0.67-0.80) and compared with Hamming, BLOSUM62 and phylogenetic baselines, with in-silico mutagenesis for escape potential.
- Still unknown: The abstract states no limitations and does not discuss generalization beyond H3N2 or beyond the data used; no wet-lab tests were run.
- AI role: A protein language model framework predicted H3N2 antigenicity from RNA sequence.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1371/journal.pcbi.1014628&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The transformer MP4 maps plain-language prompts to protein sequences. The study reports two experimentally validated de novo designs that were expressible and thermostable and showed ATP binding and hydrolysis in vitro. X-ray structures at 1.30 Å and 1.77 Å showed that one design adopts a fold not previously observed in the authors’ structural comparisons.
Technology: MP4, a transformer-based text-to-protein model
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The study reports two experimentally validated de novo designs that were expressible and thermostable and showed ATP binding and hydrolysis in vitro.
- Still unknown: Only two designed proteins are described as experimentally validated in the abstract; this is an early proof of concept, not a demonstrated general success rate. ATP binding and hydrolysis are in-vitro results, not a novel physiological function or therapeutic effect. The source describes one fold as novel; no claim is made that both structures are novel.
- AI role: MP4 converted natural-language protein specifications into candidate sequences. Researchers evaluated 96 prompts in silico, selected candidates, expressed and tested designs, and solved two crystal structures. The source attributes the designed sequences to MP4; experimental teams established expression, thermostability, ATP binding/hydrolysis and structural novelty.
Evidence sources: https://pubs.acs.org/asbcd6/article/doi/10.1021/acssynbio.6c00246/5254330/An-Artificial-Intelligence-Model-for-Translating
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Radical AI reports that AI-generated hypotheses and an autonomous lab produced new alloys, which were then tested by Purdue Applied Research Institute. During one minute under a torch, RAI-939 gained 0.25% mass while C103 lost 32%. This is a company announcement without peer review or replication details; scaling production remains to be developed.
Technology: AI-generated hypotheses and autonomous laboratory
Publication: Announcement only · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: direct test
- Tested: A head-to-head torch test run by Purdue Applied Research Institute: RAI-939 gained 0.25% mass in one minute while C103 lost 32%; the page describes no repeated trials, sample sizes or replication.
- Still unknown: Replication details are missing, the test articles were small, and production scaling is still to be developed.
- AI role: AI-generated hypotheses and an autonomous lab produced over 300 novel materials over 16 weeks.
Evidence sources: https://www.radical-ai.com/news/purdue-applied-research-institute-confirms-superior-performance-of-radical-ai-materials
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Using SINDy equation discovery on monthly observational indices from 1950–2022, researchers derived three linked low-order equations for Atlantic Multidecadal Variability, the North Atlantic Oscillation and North Atlantic precipitation. The model recovered a distinct roughly 20-year component and highlighted precipitation-linked terms as robust components of the learned dynamics; tests against the independent 1900–1939 record found skill for statistics of the variability indices and rainfall. The equations generate hypotheses about ocean-atmosphere feedbacks; they do not establish a single causal mechanism or guarantee forecasts.
Technology: Sparse Identification of Nonlinear Dynamics (SINDy) equation-discovery algorithm
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Using SINDy equation discovery on monthly observational indices from 1950–2022, researchers derived three linked low-order equations for Atlantic Multidecadal Variability, the North Atlantic Oscillation and North Atlantic precipitation.
- Still unknown: The learned equations are compact models/hypotheses from observational associations. The paper and Oxford source note that proposed mechanisms need further evaluation; reported hindcast skill is not prospective climate prediction validation.
- AI role: Searched candidate dynamical models against observational time series and selected sparse, interpretable coupled equations; the researchers analyzed robust equation terms and compared the resulting model against a historical holdout period.
Evidence sources: https://esd.copernicus.org/articles/17/1061/2026/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The Neuron abstract reports a deep-learning framework trained on trial-resolved recordings from a bidirectional human visual-cortex implant. A learned inverse network and gradient-based optimizer shaped stimulation-evoked population activity, achieved targets at lower currents and elicited more consistent perception than conventional methods. This is an experimental neural-control result, not evidence of restored functional sight. The paper was first published online on 2026-08-07; review here is limited to its abstract and metadata.
Technology: Deep learning inverse network and gradient-based optimizer for neural stimulation
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Within-study human visual-cortex stimulation experiments comparing two learned control approaches with conventional methods; lower currents and more consistent reported percepts.
- Still unknown: Complete abstract and metadata reviewed. Participant count, implant duration, long-term performance and functional-vision outcomes were not assessed; full-text XML retrieval returned HTTP 500.
- AI role: Deep learning model trained on trial-resolved neural recordings to synthesize and optimize stimulation patterns.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.1016%2Fj.neuron.2026.07.006%22&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Three-day forecasts match what earlier models achieved at two days: about 100 km trajectory error and 11 knots intensity error, compared with ECMWF-ENS and HWRF. The work is with the National Hurricane Center, CIRA, and the Met Office. The comparison uses historical 2023–2024 cyclones; why results are good at 28 km remains an open question.
Technology: WeatherNext Cyclones
Family WeatherNext: profile · models page
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Historical cyclones from 2023-2024 were benchmarked deterministically and probabilistically against ECMWF-ENS (track) and HWRF (intensity): three-day errors matched earlier two-day forecasts; a 2025 storm is cited only as an anecdote.
- Still unknown: Why the model is accurate at 28 km resolution is an open question; the 'extra day' is an average dependent on the chosen benchmarks, and the evaluation window is slightly unclear.
- AI role: A single model using Functional Generative Networks produced track, intensity and wind-structure ensembles of up to 1,000 members.
Evidence sources: https://deepmind.google/blog/weathernext-ai-model-achieves-breakthrough-in-forecasting-cyclones/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 2/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A Swarm-seq assay tested more than 500 bacterial CsrA homologs for regulation of Bacillus subtilis swarming and found two functional classes: several homologs strongly inhibited swarming while others failed despite conserved sequence and canonical GGA motifs in the hag transcript. A random-forest model trained on these assays predicted homolog activity across thousands of sequences; experimental testing supported predicted Class I regulatory activity for Bdellovibrio bacteriovorus CsrA and Pseudomonas putida RsmA. The findings indicate sequence conservation and the canonical RNA motif do not alone determine these tested regulatory effects.
Technology: Random-forest machine-learning classifier trained on Swarm-seq functional assay data
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: A Swarm-seq assay tested more than 500 bacterial CsrA homologs for regulation of Bacillus subtilis swarming and found two functional classes: several homologs strongly inhibited swarming while others failed despite conserved sequence and canonical GGA motifs in the hag transcript.
- Still unknown: Activity classifications are based on the reported B. subtilis swarming assay and tested proteins; they do not establish all native targets or regulatory functions in each source organism. The predicted function landscape across thousands of homologs is not experimentally validated wholesale. Primary publisher abstract checked; no independent replication assessed.
- AI role: High-throughput swarming assays generated a functional dataset; machine learning was then trained to classify homolog functional activity across the bacterial domain and nominate homologs for validation. Selected CsrABb and RsmAPp predictions were experimentally validated.
Evidence sources: https://academic.oup.com/nar/article/54/14/gkag756/8753004
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A genome-informed machine-learning workflow (ML-CITO) used contrastive learning over protein-language-model representations and genomic context, starting from 14 known phenazine-modifying sequences, to identify PTC. Biochemical characterization showed PTC catalyzes glutathione-dependent phenazine modification, a reaction previously assumed to occur non-enzymatically.
Technology: ML-CITO (genome-context-informed data augmentation plus protein-language-space contrastive learning)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: A genome-informed machine-learning workflow (ML-CITO) used contrastive learning over protein-language-model representations and genomic context, starting from 14 known phenazine-modifying sequences, to identify PTC.
- Still unknown: The paper reports that deleting the PTC gene did not cause a strong fitness disadvantage under tested conditions. The demonstrated chemistry and substrate context should not be generalized to other phenazine reactions.
- AI role: The model ranked candidate phenazine-interacting proteins from sparse known examples; the team then expressed and biochemically tested the candidate enzyme. AI nomination was a direct step in discovering the previously unknown enzymatic reaction.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13419639/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The XunZi AI-biologist integrated literature reasoning with multimodal molecular data and nominated CHK2 and IRAK4 kinase activation in Parkinson’s disease models. Follow-up experiments found that pharmacological or genetic Chk2 inhibition rescued dopaminergic-neuron loss and motor deficits in PD mouse models, supporting CHK2 as a disease-associated mediator in the tested systems.
Technology: XunZi AI biologist
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Follow-up experiments found that pharmacological or genetic Chk2 inhibition rescued dopaminergic-neuron loss and motor deficits in PD mouse models, supporting CHK2 as a disease-associated mediator in the tested systems.
- Still unknown: The neuroprotective evidence is from experimental Parkinson’s disease models and mice, not human patients or clinical efficacy. XunZi also nominated IRAK4, but the abstract specifically reports Chk2 inhibition rescue; do not imply equivalent causal validation for every AI-nominated target. The Nature article’s complete indexed text was read; direct open may be blocked by a cookie redirect.
- AI role: XunZi integrated logical reasoning with multimodal data fusion across biomedical literature and molecular datasets to generate and rank de novo disease-target hypotheses. Researchers then tested the CHK2 hypothesis with cellular and mouse experiments, including pharmacological/genetic inhibition; the AI nominated the target, while experiments established the reported effects.
Evidence sources: https://www.nature.com/articles/s41551-026-01769-6
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Using 164 fb⁻¹ of 13.6 TeV proton–proton data and a statistical combination with the Run 2 measurement, ATLAS reports the first observation of ZZ production with both Z bosons longitudinally polarized, at 6.5σ observed (5.6σ expected). The combined signal strength is 1.22 ± 0.21; the Run 3 measurement also reports polarization fractions across the fiducial phase space and two four-lepton invariant-mass ranges, consistent with Standard Model predictions.
Technology: Boosted decision tree (BDT) classifier
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: Using 164 fb⁻¹ of 13.6 TeV proton–proton data and a statistical combination with the Run 2 measurement, ATLAS reports the first observation of ZZ production with both Z bosons longitudinally polarized, at 6.5σ observed (5.6σ expected).
- Still unknown: The ATLAS paper was submitted to PRL as a preprint on 2026-07-14; the cited official ATLAS briefing is dated 2026-08-04. The 6.5σ first observation combines the Run 3 analysis with the prior Run 2 measurement, whose standalone significance was 4.3σ. The Run 3 analysis alone reports 5.1σ. The measured fractions agree with the Standard Model; the statistical uncertainty dominates and polarization-template modeling is the leading systematic uncertainty. This is an experimental observation from ATLAS data, not evidence of physics beyond the Standard Model.
- AI role: The BDT is trained on polarization-sensitive lepton angular observables to separate longitudinal-longitudinal ZZ events from transverse or mixed-polarization events. Its score distribution is the fitted observable used to extract the signal strength and significance.
Evidence sources: https://atlas.cern/Updates/Briefing/Polarised-Weak-Bosons
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Researchers trained models on 174 prior synthesis attempts and selected six EMT synthesis conditions for testing. Five yielded EMT by X-ray diffraction; two used starting-mixture Si/Al ratios outside the training range. EMT is an established framework, and the small, researcher-selected study does not establish a general success rate or independent replication.
Technology: XGBoost classifier, PCA and pretrained foundation models
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: As described in the primary source
- Still unknown: Independent replication and global priority not established by this review.
- AI role: XGBoost classifier, PCA and pretrained foundation models
Evidence sources: https://arxiv.org/abs/2608.03760
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Primary publisher/preprint material inspected for identity, AI role and bounded result..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
REAP combines a protein-language-model ranking predictor with active selection and automated build-and-test cycles. The study reports a P450 BM3 quintuple variant with 57-fold higher product yield than the engineered FL#62 starting scaffold in its high-throughput assay. This is a variant-specific comparison with an already engineered scaffold, not a 57-fold gain over wild type. A separate sortase result is not included in this bounded draft.
Technology: REAP; PLM-RankReg; automated build-and-test
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Publisher indexed article reports a quintuple variant and 57-fold product-yield increase over FL#62 in the high-throughput assay; kinetic characterization is also described.
- Still unknown: Independent replication and broader application are not established by this bounded review.
- AI role: Learned ranking and active selection nominate variants for automated synthesis and testing; the reported gain is measured against the engineered FL#62 scaffold.
Evidence sources: https://www.nature.com/articles/s41467-026-76264-2
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 16 Nov 2026. Review scope: Publisher indexed primary article text, title/DOI and result; PMC direct open CAPTCHA challenged. Journal/preprint chronology and sequence overlap not audited..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Zero-shot protein models prioritized HpaBC substitutions, which were experimentally assayed with a L-Dopa-responsive GFP biosensor. EVOLVEpro then combined experimentally validated substitutions, including directed-evolution-derived variants, and selected higher-order designs. The best reported combined variants produced up to 57–58 times the wild-type fluorescence signal under the stated assay conditions. This is a biosensor proxy for L-Dopa production, not a purified-enzyme kinetic measurement; the work also builds on prior directed-evolution mutations and does not establish clinical benefit.
Technology: MutCompute, ProteinMPNN, ESM protein language models and EVOLVEpro combinatorial design
Family ESM / ESMFold: profile · models page
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: 329 model-predicted variants were assayed; 19 showed at least 1.25-fold biosensor signal. EVOLVEpro-selected combined variants yielded up to 57–58-fold normalized fluorescence relative to wild type under the reported assay conditions.
- Still unknown: Fluorescence from a DopA/GFP biosensor is an indirect L-Dopa-production proxy, not a purified-enzyme catalytic constant. The work includes directed-evolution-derived mutations; G295T had prior directed-evolution support. No human therapeutic efficacy is established.
- AI role: MutCompute, ProteinMPNN and ESM prioritized single substitutions for experimental testing; EVOLVEpro combined tested mutations and iteratively selected higher-order variants, including designs incorporating directed-evolution results.
Evidence sources: https://api.biorxiv.org/details/biorxiv/10.64898/2026.07.31.742090 · https://www.biorxiv.org/content/10.64898/2026.07.31.742090v1.full
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary bioRxiv v1 API metadata and Crossref posted-date metadata checked; primary v1 full-text indexed extract reviewed for methods, biosensor-normalized outcomes, model role and affiliation. Direct full-text page open was blocked..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
OpenAI announces ten results on codes, non-sofic groups, arithmetic circuits, quantum games, lattices, and combinatorics, accompanied, according to the source, by Lean certificates. This record groups one announcement; it does not mean this collection independently verified ten results.
Technology: OpenAI internal model
Publication: Announcement only · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: prospective or independent check
- Tested: Each of the ten arguments was prepared into a manuscript by humans with the model and then formalized by the model in a Lean certificate.
- Still unknown: Not stated in the reviewed material
- AI role: An internal version of Astra found the results; humans prepared the manuscripts with the same model.
Evidence sources: https://openai.com/index/ten-advances-in-mathematics/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 5/5 · Validation 4/5 · Novelty 5/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Four classifiers (random forest, support-vector machine, logistic regression and multilayer perceptron) used six years of natural Arabidopsis leaf-microbiome data to prioritize taxa associated with Albugo infection status. Follow-up plant assays found that selected health-associated microbes reduced infection; Cystofilobasidium macerans OTU126 gave the strongest individual protection, reducing Albugo infection by an average of 73%. In a 15-member synthetic leaf community, adding this fungus increased the reported infection reduction to 40.25%, compared with 19.36% for the community alone.
Technology: Random forest, support-vector machine, logistic regression and multilayer perceptron classifiers with cross-validated feature selection
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Follow-up plant assays found that selected health-associated microbes reduced infection; Cystofilobasidium macerans OTU126 gave the strongest individual protection, reducing Albugo infection by an average of 73%.
- Still unknown: This is a controlled Arabidopsis–Albugo plant-pathogen model, not a field crop efficacy trial. ML features are associations with infection status; they do not establish the microbes caused naturally observed health status. Isolates were matched to predicted OTUs using marker sequences, which the authors note do not fully establish strain-level functional identity. Genomic terpene-cluster hypotheses were not directly tested. No commercial or field-ready biocontrol performance is established.
- AI role: Models classified infected versus uninfected plants from microbial community profiles and prioritized disease-associated and health-associated taxa for follow-up. Researchers then tested selected cultured isolates in planta and tested Cystofilobasidium in a synthetic community.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/42521687/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The off-nuclear transient TDE 2025abcr revealed a dormant supermassive black hole about 9.3 kiloparsecs from its host galaxy’s center, with a mass described as similar to the Milky Way’s central black hole. A machine-learning program searching for tidal-disruption flares anywhere in the sky selected the event for follow-up three months after deployment; astronomers then collected multi-instrument evidence and interpreted it as the strongest known wandering-black-hole case. Its precise origin remains uncertain, with stripped remnant and black-hole-ejection scenarios both discussed.
Technology: Off-nuclear tidal-disruption-event classifier (tdescore)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The off-nuclear transient TDE 2025abcr revealed a dormant supermassive black hole about 9.3 kiloparsecs from its host galaxy’s center, with a mass described as similar to the Milky Way’s central black hole.
- Still unknown: The specific off-nuclear transient was AI-selected, but the paper’s evidence and interpretation are observational follow-up, not a model-generated discovery. The black hole’s formation history is unresolved; the host has no obvious surrounding galaxy at the black-hole position.
- AI role: Ranked the transient as a candidate tidal-disruption event away from a galaxy nucleus, prompting spectroscopic and multi-wavelength follow-up; the astrophysical interpretation relied on subsequent human analysis and observations.
Evidence sources: https://doi.org/10.3847/2041-8213/ae77f3
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A reinforcement-learning agent tuned over 1,000 control parameters on the Willow chip: logical stability was 3.5 times higher under injected drift, with a further 20% reduction in logical error after expert calibration, and error below one per thousand cycles in the surface code. Drift was injected artificially, and scalability was shown only in simulation; published in Nature.
Technology: Reinforcement learning for quantum control
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: On the Willow chip, drift was injected artificially into control parameters; RL steering raised logical stability 3.5-fold, and a further 20% error reduction followed expert calibration; scalability was tested only in simulation.
- Still unknown: The drift was artificial rather than natural; scalability evidence is simulation-only; environmental decoherence cannot be fully prevented; fuller results need tighter integration.
- AI role: An RL agent used error-detection events as a learning signal and adjusted frequencies, amplitudes and phases during the computation without stopping it for recalibration.
Evidence sources: https://research.google/blog/towards-a-quantum-computer-that-learns-from-its-errors/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Confounder-aware machine-learning feature selection on plasma proteomics identified a seven-protein signature—AHSG, CLEC3B, COMP, F9, LRG1, MCAM and MRC1—to classify remission versus active ANCA-associated vasculitis (AAV). After targeted PRM mass-spectrometry assay development, a generalized linear model was evaluated in an independent cohort of 108 AAV patients and achieved AUC 0.94 (95% CI 0.90–0.98); combined discovery and validation performance was AUC 0.97. In the study cohorts, the panel outperformed routine CRP and ANCA titer measurements.
Technology: MetaDeconfoundR confounder-aware feature analysis, LASSO and Boruta feature selection, and a generalized linear model applied to targeted PRM mass-spectrometry data
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: After targeted PRM mass-spectrometry assay development, a generalized linear model was evaluated in an independent cohort of 108 AAV patients and achieved AUC 0.94 (95% CI 0.90–0.98); combined discovery and validation performance was AUC 0.97.
- Still unknown: This is cross-sectional classification of disease state in patients already diagnosed with AAV, not a prospective prediction of relapse, a validated basis for stopping immunosuppression, or evidence of improved clinical outcomes. The two cohorts together included just over 200 patients with a rare disease; the authors note uncertainty from sample size and treatment heterogeneity, and the panel remains a starting point for clinical-tool development. It is an adjunct biomarker result, not proof that the seven proteins cause remission or vasculitis activity.
- AI role: Machine-learning methods accounted for clinical covariates including CRP, kidney function and ANCA-related measures, prioritized candidate plasma proteins, and reduced these to a seven-protein panel. Targeted PRM-MS quantified the selected proteins, and the GLM classified AAV remission versus active disease in an independent cohort. The measured validation performance and protein signature are the specific AI-enabled diagnostic finding.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/42481524/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A supervised DFT-trained model and Gaussian-process experimental loop guided selection of BaCo0.8Nb0.1Zr0.1O3−δ. The publisher abstract reports 2.68 W cm−2 peak fuel-cell power at 600 °C and a separate 500-hour electrolysis test. These are specified laboratory cell tests; the Research Square v1 and journal versions differ in candidate counts and are not conflated here.
Technology: DFT-trained supervised model; Gaussian-process Bayesian optimization
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Publisher primary abstract reports BCNZ20 and specified fuel-cell and electrolysis measurements.
- Still unknown: Version-specific candidate counts and detailed test protocols require full-text comparison; no commercial performance claim.
- AI role: Supervised composition/property prediction narrows candidates; Bayesian experimental optimization iteratively guides measurement.
Evidence sources: https://doi.org/10.1002/adma.74212
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Primary source checked for identity, publication date, AI workflow and the bounded experimental result stated..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Researchers combined experimental activity measurements with machine learning to enrich a second MthUPO peroxygenase library for substrate-specific activity. Across 506 distinct variants screened and sequenced, 97 of 108 variants in the ML-enriched library outperformed wild type in the screening assay, compared with 14% in the initial htFuncLib library. The authors report an 11.8-fold turnover number for the best variant producing 4-hydroxy-beta-damascone, with greater than 99% regioisomeric excess. Separate 5–12-fold enrichment and recovery figures in the paper are simulation benchmarks and are not treated as experimental outcomes.
Technology: Machine-learning enrichment of experimental enzyme-activity data following htFuncLib design
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: The preprint reports screening and sequencing 506 distinct MthUPO variants. In the ML-enriched library, 97 of 108 variants outperformed wild type in the screening assay, compared with 14% in the initial htFuncLib library. The best reported variant had an 11.8-fold turnover number for 4-hydroxy-beta-damascone with greater than 99% regioisomeric excess.
- Still unknown: The preprint is not peer reviewed. The paper separately reports simulation-only MLEE/htFuncLib comparisons, which are excluded from the experimental result summary. Assay outcomes do not establish process-scale productivity or commercial suitability.
- AI role: Machine learning used experimental enzyme-activity measurements to enrich a library of MthUPO variants for testing, following an earlier htFuncLib design stage.
Evidence sources: https://api.biorxiv.org/details/biorxiv/10.64898/2026.07.20.739108 · https://www.biorxiv.org/content/10.64898/2026.07.20.739108v1.full
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Official bioRxiv API v1 metadata/abstract and primary indexed v1 full-text excerpts; simulation and experimental results kept separate..
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A six-cohort plasma proteomics study used cross-validated LASSO to select a 21-protein multiplex panel for dementia differential diagnosis. In the new DLB/FTD results, ITGAV and ITGAM were reduced in dementia with Lewy bodies and also tracked Lewy-body pathology in independent autopsy-confirmed cohorts; a panel including ITGAV, ITGAM and ERBB3 distinguished DLB from controls and Alzheimer dementia in an independent multicenter cohort. For frontotemporal dementia, NfL and B4GAT1 contributed to the panel distinguishing FTD from controls. Validation-set AUCs were 0.83 for DLB versus controls, 0.78 for DLB versus Alzheimer dementia, 0.85 for FTD versus controls and 0.77 for FTD versus Alzheimer dementia.
Technology: Cross-validated LASSO regression and multivariable logistic-regression models applied to Olink proximity-extension plasma proteomics
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A six-cohort plasma proteomics study used cross-validated LASSO to select a 21-protein multiplex panel for dementia differential diagnosis.
- Still unknown: These are observational plasma biomarker associations and retrospective diagnostic-validation results, not evidence that the proteins cause disease or improve patient outcomes. The work reports moderate rather than definitive discrimination, especially DLB versus Alzheimer dementia (validation AUC 0.78) and FTD versus Alzheimer dementia (0.77); external prospective utility remains unestablished. The later study reuses the bPRIDE cohorts and includes Alzheimer-disease analyses overlapping the group’s earlier 2025 bPRIDE plasma-proteomics paper; this entry is limited to the distinct DLB/FTD differential-marker and independent 21-plex validation findings. No individual biomarker is presented as a stand-alone clinical test.
- AI role: The analysis used 100-fold cross-validated LASSO feature selection to prioritize a limited set of disease-associated proteins and choose 21 markers for a custom multiplex assay. Pairwise logistic-regression models used selected markers to distinguish DLB and FTD from controls and Alzheimer dementia; the targeted panel was then measured in an independent multicenter validation cohort.
Evidence sources: https://www.nature.com/articles/s43587-026-01162-7
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A research project used ChatGPT 5.5 Pro and Codex agents to develop proof ideas and proofs for five Banach-space problems. The resulting paper gives affirmative solutions to questions on toroidal Elton–Odell separation, non-Calkin unital Banach algebras, strict cosingularity and adjoints, weakly compact basis factorization, and primariness of Lp(L1). The authors report that models generated the core mathematics; human mathematicians checked and refined the proofs. The results are presented in an unrefereed preprint, and the authors take responsibility for correctness.
Technology: ChatGPT 5.5 Pro and Codex agents using an automated open-problem search pipeline
Family Codex: profile · models page
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: A research project used ChatGPT 5.5 Pro and Codex agents to develop proof ideas and proofs for five Banach-space problems.
- Still unknown: Unrefereed arXiv preprint. Authors state that humans verified and refined the proofs and are responsible for correctness; the proofs are not formally machine-checked. The five problems were selected by mathematicians, while a separate pipeline component automatically found other candidate problems. The five results remain bundled as one paper-level report.
- AI role: In the pipeline component, agents searched arXiv literature for open problems and attempted solutions. For the five selected Banach-space problems, the authors say model attempts generated the proof ideas and essentially complete arguments before human verification, editing and integration; models were not limited to exposition.
Evidence sources: https://arxiv.org/abs/2607.17388
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
By combining single-cell splicing-QTL analysis during influenza A infection with SpliceAI/Pangolin annotation and statistical fine-mapping, researchers prioritized causal splice variants. They experimentally characterized rs2297616 in PARP2: its A allele creates an alternative 5′ splice donor, adding 39 nucleotides and producing a protein isoform 13 amino acids longer. CRISPR-edited A549 cells showed the predicted splice and protein changes; the A-allele edited cells also had higher influenza A infection than G-allele cells. The variant was associated with higher viral burden in a panel of 96 lymphoblastoid cell lines.
Technology: SpliceAI and Pangolin deep-learning splice-effect predictors, integrated with single-cell scHi-HOST sQTL mapping and SuSiE statistical fine-mapping
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: They experimentally characterized rs2297616 in PARP2: its A allele creates an alternative 5′ splice donor, adding 39 nucleotides and producing a protein isoform 13 amino acids longer.
- Still unknown: The 96-donor LCL association is observational and modest (the paper reports p=0.029 for the genotype association). Infection experiments used edited A549 lung epithelial cells, not patients or an in-vivo infection model. The same locus had been linked to alternative splicing previously, but the paper provides new evidence connecting genotype to PARP2 protein isoform expression and to influenza susceptibility. Broader viral-susceptibility implications remain unproven.
- AI role: Deep-learning splice models prioritized variants predicted to alter donor/acceptor sites; statistical fine-mapping and sQTL association narrowed causal candidates. The AI-supported analysis nominated rs2297616; the authors then tested its splice and protein-isoform effects using reporter assays, CRISPR-edited A549 cells, sequencing and western blotting, and assessed influenza infection in edited cells.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13663244/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
EnzGFM-Agent scored all possible single amino-acid substitutions in secreted alkaline phosphatase (SEAP) and selected top candidates from EnzGFM-1.5B and ESM2-3B for testing in transfected Chinese hamster ovary cells. Among the five variants uniquely prioritized by EnzGFM, all five had significantly higher measured activity than wild type in the reported assay; three exceeded 1.5-fold, and A46I reached 6.52-fold. This is an experimentally measured enzyme-engineering result for the tested SEAP variants.
Technology: EnzGFM-Agent with EnzGFM-1.5B and ESM2-3B
Family ESM / ESMFold: profile · models page
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: EnzGFM-Agent scored all possible single amino-acid substitutions in secreted alkaline phosphatase (SEAP) and selected top candidates from EnzGFM-1.5B and ESM2-3B for testing in transfected Chinese hamster ovary cells.
- Still unknown: The evidence is limited to single substitutions in one SEAP reporter protein tested after transfection in CHO-S cells, with three independent biological replicates; it does not establish improved activity for other enzymes, industrial process performance, or a naturally occurring enzyme function. The five EnzGFM-specific variants all differed significantly from wild type, but the study also evaluated ESM2-ranked variants and five variants shared by both models. This is a paper-specific experimental validation, not independent replication.
- AI role: The workflow scored every possible SEAP single-point substitution, ranked candidates using EnzGFM-1.5B and ESM2-3B, and selected 15 unique variants for experimental validation. The distinct EnzGFM-prioritized group comprised five variants; the paper reports significantly increased SEAP activity for all five, with A46I at 6.52-fold wild type.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13494029/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Analysis of the CNN-detected ComPACT Sunyaev–Zel’dovich catalogue identified 116 previously unreported galaxy clusters with new redshift estimates, including five massive systems (M500c > 6 × 10^14 solar masses) at z > 0.7. The authors report 158 new cluster-mass determinations and higher completeness than ACT DR5 at fixed purity, especially for massive high-redshift systems.
Technology: ComPACT deep-learning convolutional neural network catalogue, followed by optical photometric redshift and SZ mass analysis
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Analysis of the CNN-detected ComPACT Sunyaev–Zel’dovich catalogue identified 116 previously unreported galaxy clusters with new redshift estimates, including five massive systems (M500c > 6 × 10^14 solar masses) at z > 0.7.
- Still unknown: The ComPACT catalogue contains 2,962 SZ-selected candidates; the paper confirms approximately 60% as clusters, so the full candidate list must not be treated as confirmed. Its newly reported cluster count relies on cross-matches and optical photometric-redshift significance criteria with an estimated 5% false-positive rate; most are not presented as new spectroscopic confirmations. Masses are scaling-relation estimates. The five high-mass, high-redshift systems are identified with redshift algorithms and SZ measurements, and their interpretation remains subject to catalogue selection/completeness and mass-estimation uncertainties.
- AI role: The CNN detected SZ cluster candidates in combined ACT+Planck maps, creating the candidate catalogue analyzed here. The authors used existing catalogues and DESI Legacy Imaging Survey photometry to validate systems and estimate redshifts, and ACT+Planck/Planck Compton-y maps with SZ scaling relations to derive masses. The CNN catalogue has higher completeness than ACT DR5 at fixed purity, particularly at high mass and redshift; it is the entry point enabling these new cluster identifications and measurements.
Evidence sources: https://www.aanda.org/articles/aa/full_html/2026/07/aa59201-26/aa59201-26.html
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A deep-learning model trained on deep-mutational-scanning sequence–binding data selected minimal mutation combinations for a human anti-PD-L1 antibody with mouse cross-reactivity. Variants with four to five substitutions showed in-vitro and cellular binding comparable to conventionally recombined variants with 13–15 substitutions; the paper also reports blockade assays and therapeutic activity in a mouse tumor model. These are preclinical antibody-engineering results, not evidence of human efficacy. The preprint reports patent applications related to the subject matter.
Technology: Deep-learning model trained on deep-mutational-scanning sequence–binding data
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: The preprint reports in-vitro and cellular binding properties of selected variants, PD-1/PD-L1 competition/blockade assays, epitope mapping and mouse tumor-model assessment. The 4–5 substitution variants are compared with highly mutated variants requiring 13–15 substitutions.
- Still unknown: Mouse-model activity is preclinical and does not establish human therapeutic efficacy. The preprint reports patent applications relevant to the work; no freedom-to-operate or patent-priority conclusion is made.
- AI role: A deep-learning model trained on DMS-derived sequence-binding data selected minimal mutation combinations for experimental antibody characterization.
Evidence sources: https://www.biorxiv.org/content/10.64898/2026.07.15.738714v1.full · https://www.biorxiv.org/content/10.64898/2026.07.15.738714v1 · https://api.biorxiv.org/details/biorxiv/10.64898/2026.07.15.738714
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary bioRxiv v1 indexed full-text/abstract, author-affiliation metadata and official API metadata; direct page open returned 403..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
UniKP was used to rank sulfotransferase homologues by predicted kinetics. The study reports experimental testing of five prioritized candidates and identifies EfST1A1, with about 6.5-fold higher in-vitro activity than the initial hit under the reported assay. Later directed evolution is not attributed to the model. This is a bounded enzyme-discovery result, not proof of industrial production or broad enzyme generality.
Technology: UniKP kinetic prediction
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Publisher abstract/indexed article text reports five tested AI-ranked enzymes and identifies EfST1A1; institute research announcement summarizes experimental activity and later evolution.
- Still unknown: Independent replication and broader application are not established by this bounded review.
- AI role: UniKP ranks enzyme homologues by predicted kinetic properties; researchers tested top candidates and separately applied directed evolution to the selected hit.
Evidence sources: https://www.sciencedirect.com/science/article/abs/pii/S1385894726047108
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 16 Nov 2026. Review scope: Publisher article preview/indexed text and institute source; full publisher text inaccessible. Issue date is recorded; exact online-first and earlier version not established..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A PCA-enhanced Transformer classifier screened 298,195 SDSS DR7 spectra and selected 14 TDE-like cases for secondary review. The authors report two newly identified TDEs: SDSS J124225.39+642919.0, whose GALEX UV source was present by the spectrum date (before 11 February 2002), and SDSS J152459.70+045423.1, whose TDE-H+He spectrum coincided with a Catalina optical outburst in 2007–08. The first is reported as the earliest optical–UV TDE then known; one other model-selected source was a previously reported likely TDE.
Technology: PCA-enhanced Transformer TDE spectrum classifier
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: A PCA-enhanced Transformer classifier screened 298,195 SDSS DR7 spectra and selected 14 TDE-like cases for secondary review.
- Still unknown: The source is an arXiv preprint whose arXiv record says accepted by The Astrophysical Journal; no journal DOI was supplied in the source checked. The TDE classifications are supported by spectral and contemporaneous archival photometric evidence, not new contemporaneous follow-up observations. The event dates predate 2020, but this record uses the 2026 source-publication date for the since-2020 catalog period. The paper also reports a third previously known likely TDE, which is not counted as a new discovery.
- AI role: The Transformer classifier screened preprocessed SDSS galaxy spectra and selected TDE-like events for follow-up assessment. From roughly 298,000 spectra it produced 14 cases above the selection threshold; after checking spectral features against archival GALEX or Catalina variability and other diagnostics, the authors reported two previously unreported TDEs and one previously reported likely TDE. Reported test-set precision and recall were 0.88 and 0.99, respectively.
Evidence sources: https://arxiv.org/abs/2607.11539
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A Bayesian-optimization-guided mechanochemical study identified conditions for the established Johnson–Corey–Chaykovsky reaction using KOH in place of hazardous NaH, under air-equilibrated conditions and with no solvent or liquid-assisted grinding agent. In the chalcone benchmark, one optimized condition reached cyclopropanation completion in five minutes; the study also demonstrated cyclopropanation and epoxidation across multiple substrates, including yields of 45–75% for four additional substrates.
Technology: Bayesian optimization implemented with BoTorch/PyTorch, including multi-objective optimization of product yield and process mass intensity
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: A Bayesian-optimization-guided mechanochemical study identified conditions for the established Johnson–Corey–Chaykovsky reaction using KOH in place of hazardous NaH, under air-equilibrated conditions and with no solvent or liquid-assisted grinding agent.
- Still unknown: The Johnson–Corey–Chaykovsky transformation itself is established chemistry; the finding is a measured, safer and faster mechanochemical process window rather than discovery of a new reaction class. A 5-minute completion time and other yields apply to the tested substrates/conditions; some prior solvent-minimized conditions could not be reproduced. The scale and substrate scope remain those reported in the paper.
- AI role: The BO models selected among six reaction variables (base identity/loading, time, milling frequency, liquid-assisted grinding agent and loading) using experimental yields and process-mass-intensity outcomes. Researchers ran the experiments, performed synthetic substrate-scope tests and used time-resolved in-situ X-ray diffraction to study milling effects.
Evidence sources: https://www.nature.com/articles/s41467-026-75499-3
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A multiscale microscopy study of carbonaceous chondrite WIS 91600 used a deep-learning classifier to segment magnetic remanence-carrying ensembles and quantify their volume and texture. Combined chemical mapping, 3D nanotomography, and diffraction showed fine matrix-hosted magnetite has paleomagnetic recording potential comparable to magnetite framboids but is less susceptible to strong interactions; transient heating also reduced the contribution of matrix pyrrhotite through its vacancy-ordering state. This is one meteorite case study, and the mineralogical conclusions draw on the full microscopy and diffraction workflow, not AI alone.
Technology: UNET segmentation, SIGMA interactive machine-learning mineral analysis, and few-shot image segmentation
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: A multiscale microscopy study of carbonaceous chondrite WIS 91600 used a deep-learning classifier to segment magnetic remanence-carrying ensembles and quantify their volume and texture.
- Still unknown: The source is a single meteorite case study. The AI classified/quantified image ensembles, while mineral identities and remanence interpretation use additional microscopy, chemical mapping, nanotomography, and diffraction; extrapolation to other carbonaceous chondrites requires more samples.
- AI role: Classified remanence-carrying ensembles in microscopy images and quantified their volumes and morphologies; complementary imaging and diffraction supported mineral identification and interpretation.
Evidence sources: https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2025GC012882
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Trained with a loss function that preserves energy across scales, it reaches a mean squared error equal to the Met Office global model and exceeds it on some metrics, with more realistic storm cores in Hurricane Ian and Storm Ciarán. The announcement gives no figures, covers only two extreme cases, and the system is not operational.
Technology: FastNet (spherical-harmonic loss)
Family FastNet: profile · models page
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: FastNet was compared with the Met Office global model on verification measures (RMSE and some other metrics, shown only in a figure) and qualitatively on Hurricane Ian and Storm Ciaran storm cores.
- Still unknown: The model is not operational, the text gives no numerical results, only two extreme cases are described, and comparable skill does not establish the same validation as the Global Model.
- AI role: A machine-learning weather model co-developed with the Alan Turing Institute, trained with a modified spherical-harmonic loss to keep energy across scales physically realistic.
Evidence sources: https://www.metoffice.gov.uk/about-us/news-and-media/media-centre/weather-and-climate-news/2026/new-research-shows-ai-can-produce-trustworthy-physically-realistic-weather-forecasts-
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A computational histology AI platform built a prognostic signature from diagnostic whole-slide images in 507 CHAARTED participants and evaluated the locked signature in 684 ENZAMET participants. In the independent ENZAMET cohort, the unfavorable-risk group had worse overall survival (HR 2.6, 95% CI 2.0–3.4) and progression-free survival (HR 2.2, 95% CI 1.7–2.7); associations remained after adjustment for clinical variables. The article also reports treatment-interaction analyses, but this retrospective secondary analysis does not establish prospective treatment-selection benefit.
Technology: Computational Histology Artificial Intelligence (CHAI) platform
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: In the independent ENZAMET cohort, the unfavorable-risk group had worse overall survival (HR 2.6, 95% CI 2.0–3.4) and progression-free survival (HR 2.2, 95% CI 1.7–2.7); associations remained after adjustment for clinical variables.
- Still unknown: The locked image-only biomarker was independently evaluated across two prospective randomized-trial datasets, but this article is a post hoc secondary analysis, not a prospective biomarker-directed treatment trial. The treatment-interaction result is suggestive and does not prove that choosing triplet therapy by this score improves outcomes. No clinical deployment or benefit claim is made. Primary journal abstract checked; no independent replication assessed.
- AI role: CHAI extracted quantitative histomorphologic features from H&E diagnostic whole-slide images and learned an image-only risk signature on CHAARTED data; the locked signature was tested in ENZAMET. Humans specified cohorts/endpoints and analyzed the trial outcomes.
Evidence sources: https://ascopubs.org/doi/10.1200/PO-26-00321
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
ARCADIAMP combines iterative discrete diffusion generation, ESM2-based activity classifiers and sequence novelty filtering. Eight of ten experimentally screened candidates met the paper’s MIC threshold; Arcinin showed activity against tested pathogens and reduced bacterial burden in a mouse wound infection model. These are selected preclinical findings, not evidence of human safety or treatment efficacy; formal resistance-evolution testing remains future work.
Technology: Iterative discrete diffusion model with ESM2 classifiers and novelty filter
Family ESM / ESMFold: profile · models page
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: As described in the primary source
- Still unknown: Independent replication and global priority not established by this review.
- AI role: Iterative discrete diffusion model with ESM2 classifiers and novelty filter
Evidence sources: https://www.nature.com/articles/s41467-026-75030-8
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Primary publisher/preprint material inspected for identity, AI role and bounded result..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The Quake Neural Operator detects and locates microseismic events directly from continuous data without explicit phase picking. Applied to the Geysers geothermal field, it finds nearly an order of magnitude more seismic detections than routine catalogues, checked against the Phase Neural Operator and visual inspection rather than an independent ground-truth catalogue.
Technology: Quake Neural Operator (QNO), multi-task neural operator for classification and regression
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Detections at the Geysers geothermal field were compared with the Phase Neural Operator and visual inspection, and the method was reported to find nearly an order of magnitude more detections than routine catalogs.
- Still unknown: Not reported; the detections are checked only against the Phase Neural Operator and visual inspection.
- AI role: Quake Neural Operator detects and locates microseismic events directly from continuous data without explicit phase picking.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1038/s41467-026-73965-6&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A genome-wide CRISPR knockout screen in primary human epidermal keratinocytes, combined with VirtualCRISPR language-model prioritization, identified ALOX5 and OXTR as previously underappreciated cell-intrinsic regulators of surface IL17RA. Targeted knockout in keratinocytes from two adult donors reduced IL17RA and inflammatory cytokine secretion. Topical inhibition of ALOX5 or OXTR also reduced disease signs in an imiquimod mouse model of psoriasiform dermatitis.
Technology: VirtualCRISPR, a functional-genomics language-model framework trained on BioGRID-ORCS screens
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: A genome-wide CRISPR knockout screen in primary human epidermal keratinocytes, combined with VirtualCRISPR language-model prioritization, identified ALOX5 and OXTR as previously underappreciated cell-intrinsic regulators of surface IL17RA.
- Still unknown: The genetic validation used primary cells from two donors, while therapeutic effects were tested in an imiquimod-induced mouse model, not patients. The screen provided the initial unbiased hits; the AI supplied an orthogonal novelty-prioritization layer and did not itself measure biological effects. The paper reports a provisional patent application covering methods and findings. No clinical efficacy, safety or topical treatment benefit in humans is established.
- AI role: After the unbiased genome-wide CRISPR screen, VirtualCRISPR scored whether perturbing each gene in primary adult keratinocytes would alter IL17RA expression. The authors combined screen enrichment with the model’s lower prior probability for noncanonical hits to prioritize high-novelty regulators; targeted CRISPR knockout and pharmacologic mouse experiments tested the resulting biology.
Evidence sources: https://www.nature.com/articles/s41467-026-75249-5
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Using Euclid Wide Survey photometry over about 3,000 square degrees, the team applied several photometric selection methods, including XGBoost classifiers and probabilistic modeling, then obtained telescope spectra for candidates. Spectroscopy confirmed 31 new quasars at redshifts 6.6–7.8; 12 are at z≥7, and the paper reports EUCL J172902.75+641018.1 at z≈7.77 as the highest-redshift quasar known at publication. The sample extends quasar studies to fainter luminosities at z>7.
Technology: XGBoost machine-learning classifiers and Extreme-Deconvolution high-redshift quasar selection, alongside probabilistic spectral-energy-distribution methods
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Using Euclid Wide Survey photometry over about 3,000 square degrees, the team applied several photometric selection methods, including XGBoost classifiers and probabilistic modeling, then obtained telescope spectra for candidates.
- Still unknown: Peer-reviewed Euclid Consortium study. Candidate ranking was AI-assisted, but human review and spectroscopic follow-up were essential to confirmation. The redshift record is the authors’ claim at the article’s publication date, not a claim about the present-day record. This is one paper-level sample of 31 confirmed quasars; candidate lists and the 31 individual objects are not separate entries.
- AI role: Multiple photometric selection methods ranked high-redshift quasar candidates from Euclid colors; XGBoost estimated quasar probabilities and redshifts. Humans visually removed artifacts and prioritized follow-up, and Keck, Magellan and LBT spectra established the 31 quasar identifications.
Evidence sources: https://doi.org/10.1051/0004-6361/202658883
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A closed-loop active-learning campaign for Borrelia burgdorferi whole-cell screening increased the experimental hit rate five-fold, from 0.2% in investigator-selected plates to 1.0% in the AL-guided campaign. The same study separately reports prospective in-silico selection of diverse compounds across bacterial species, with 11.0% experimental validation of predicted hits and 53-fold enrichment over investigator-directed screening; those validated compounds showed the intended narrow-spectrum B. burgdorferi activity. These are assay and screening results, not clinical efficacy.
Technology: Active learning using a directed message-passing neural network and evidential Dirichlet head
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: In a closed-loop Borrelia burgdorferi campaign, active-learning-guided selection produced a 1.0% experimental screening hit rate versus 0.2% in investigator-selected plates. A separate prospective in-silico selection across bacterial species achieved 11.0% experimental validation of predicted hits (53-fold enrichment over investigator-directed screening); the paper reports intended narrow-spectrum activity for B. burgdorferi.
- Still unknown: Preprint evidence is not independent replication or clinical testing. The cross-species hit-validation experiment is distinct from the closed-loop Borrelia hit-rate comparison. Direct full page retrieval was blocked; checked material was the primary indexed preprint and archived primary article text.
- AI role: A directed message-passing neural network with an evidential Dirichlet head supported active-learning compound selection for whole-cell inhibition testing; selected compounds were then experimentally screened.
Evidence sources: https://www.biorxiv.org/content/10.64898/2026.07.04.736489v1 · https://pmc.ncbi.nlm.nih.gov/articles/PMC13344992/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary bioRxiv indexed text, PubMed metadata/abstract and PMC indexed full-text excerpt; direct bioRxiv and PMC page loads were blocked..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A fine-tuned sequence model and generative workflow guided identification of xylanase 8E20 and mutant 8E20-178. The primary abstract reports an optimum-pH shift from 8 to 10 and a 1.9-fold improvement, but calls the metric catalytic activity while indexed conclusion text calls it catalytic efficiency; this draft preserves the ambiguity. It does not infer unreported assay conditions or industrial-scale performance.
Technology: AAEPre sequence model; xynODM generative model
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Primary abstract reports 8E20-178, pH optimum shift and 1.9-fold improvement.
- Still unknown: Publisher full text inaccessible; exact improvement metric and assay conditions unresolved. Prior sequence characterization and independent replication not assessed.
- AI role: Sequence models and composite selection index prioritized candidate enzyme; generative model guided variant engineering.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/42398548/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Primary source checked for identity, publication date, AI workflow and the bounded experimental result stated..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Giorgio Parisi and Francesco Zamponi provide an analytic proof that the critical exponents a and b in the full replica-symmetry-breaking scaling solution for infinite-dimensional dense hard spheres obey a+b=1. The identity links gap, contact-force and overlap scaling relations and had previously been known numerically but not analytically. The authors report that Claude Opus 4.7 derived the core proof idea with minimal supervision, with Sonnet 4.6 helping refine it; the authors found and corrected errors in an early draft, then verified and edited the final proof.
Technology: Anthropic Claude Opus 4.7 and Claude Sonnet 4.6
Family Claude: profile · models page
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Giorgio Parisi and Francesco Zamponi provide an analytic proof that the critical exponents a and b in the full replica-symmetry-breaking scaling solution for infinite-dimensional dense hard spheres obey a+b=1.
- Still unknown: The theorem is scoped to the fullRSB scaling equations for the infinite-dimensional dense hard-sphere jamming solution; it is not a general proof about every finite-dimensional jamming system. The result is peer-reviewed and published, but proof validation and correction are attributed to the human authors.
- AI role: After numerical investigation, Opus generated the analytic proof strategy and proof draft for the previously unproven scaling identity; Sonnet refined minor steps. The authors report 40 prompts, corrected early inconsistencies and verified/edited the published proof.
Evidence sources: https://doi.org/10.1088/1742-5468/ae7bd7
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
An HMM and supervised PET hydrolase model screened sequence candidates and prioritized the PET-binding module tCBM13-1. Experiments found greater PET adsorption and retention than the CBM2 comparator; when displayed with FAST-PETase on E. coli, the system increased PET-film depolymerization by about 43% and retained over 64% of initial activity after ten cycles. These are laboratory assays, not evidence of industrial-scale recycling.
Technology: Profile-HMM homology search combined with supervised PET hydrolase machine-learning model PETML
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The authors experimentally measured tCBM13-1 adsorption and retention on PET and tested whole-cell co-display with FAST-PETase. The publisher abstract reports about 43% greater PET-film depolymerization (2793 μg/(d·cm²)) and over 64% residual activity after ten cycles.
- Still unknown: Results are laboratory PET assays; they do not establish industrial-scale recycling, environmental deployment, or lifecycle benefit.
- AI role: A supervised PET hydrolase machine-learning model within an HMM-based sequence-mining workflow prioritized tCBM13-1 from candidate modules for experimental adsorption and whole-cell depolymerization tests.
Evidence sources: https://www.sciencedirect.com/science/article/pii/S0960852426013611 · https://pubmed.ncbi.nlm.nih.gov/42385911/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary publisher-indexed full text and primary NCBI/PubMed metadata/abstract, including official article date and affiliations..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Audio embeddings from the BirdNET model robustly distinguished individual North Atlantic right whales in upcall recordings from 234 samples across 11 individuals at 3 sites. Simulations showed that an extra 7 dB or more of signal excess is needed to identify individuals under ambient noise. The study uses on-animal archival tags on an endangered species, but the sample is small and the model was not trained for whales.
Technology: BirdNET audio embeddings (deep learning)
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: BirdNET audio embeddings were tested for distinguishing 11 individual right whales in 234 upcall samples, and noise simulations estimated the extra signal excess needed for individual identification compared with detection.
- Still unknown: The abstract lists no explicit limitations; noise effects are simulated rather than field-measured, and the dataset is small (11 individuals at 3 sites).
- AI role: The BirdNET deep-learning model generated audio embeddings used to classify individual identity.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1371/journal.pcbi.1013321&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A two-stage deep-learning pipeline searched LAMOST DR12 spectra and retained 27,853 BHB candidate spectra. After removal of previously known objects, stringent Balmer-line-profile fitting, and deduplication, the authors report 3,583 unique newly identified blue horizontal-branch stars. Gaia photometry provides an additional consistency check.
Technology: MSFA-Net, a multi-scale and frequency-attention deep neural network, followed by a binary refinement classifier
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: A two-stage deep-learning pipeline searched LAMOST DR12 spectra and retained 27,853 BHB candidate spectra.
- Still unknown: This is an arXiv preprint. The paper says high-confidence objects were identified by Balmer-line profile fitting and Gaia color-magnitude consistency, while acknowledging that a fraction could still be contaminants; some atmospheric-parameter outliers are attributed to inference systematics. Counts are unique objects after deduplication, not 3,923 spectra.
- AI role: MSFA-Net searched the LAMOST DR12 hot-star spectral domain, used staged classification to reduce the spectra to candidate sets, and aided construction of the new sample; Balmer-line profile criteria and duplicate removal yielded the reported 3,583 unique objects. A separate SLAM model estimated atmospheric parameters rather than serving as the confirmation step.
Evidence sources: https://arxiv.org/abs/2606.29918
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The HOLISMOKES search identified the ZTF transient SN 2025wny at the position of the previously selected lens candidate PS1 J0716+3821. Follow-up imaging resolved multiple supernova images and spectroscopy classified it as a Type I superluminous supernova at z=2.008 ± 0.001; the second deflector galaxy was measured spectroscopically at z=0.375 ± 0.001. This is the first confirmed galaxy-scale strongly lensed supernova with expected time delays of days to weeks, offering a new system for time-delay cosmography.
Technology: Deep-learning Pan-STARRS 3π lens-candidate classification and transient-alert cross-matching pipeline
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The HOLISMOKES search identified the ZTF transient SN 2025wny at the position of the previously selected lens candidate PS1 J0716+3821.
- Still unknown: The redshift, SN type and lensing are based on follow-up observations; time delays are predicted from image separations and were still awaiting high-cadence measurement in this paper. The authors note that the SN’s brightness and resolvability introduce selection effects. An independent paper analyzed the same transient; the discovery is counted once here, not as separate detections by each follow-up team.
- AI role: The discovery followed a workflow that first classified all 3 billion Pan-STARRS 3π sources with deep learning to select high-quality strong-lens candidates, then continuously cross-matched those positions to ZTF alerts. The team matched the 2025 transient to the preselected lens candidate and triggered imaging and spectroscopic follow-up, which established the lensed SLSN result.
Evidence sources: https://www.aanda.org/articles/aa/full_html/2026/06/aa57847-25/aa57847-25.html
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Machine-learning ink detection on synchrotron scans let papyrologists read about 22 columns of a likely Stoic ethics text from scroll PHerc. 1667. The text is fragmentary because only the inner core survives, and humans transcribed and reviewed every reading.
Technology: Machine-learning ink detection with virtual unwrapping
Publication: Preprint · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: prospective or independent check
- Tested: Readings checked against explicit coverage and papyrological-review criteria; ink recovery validated independently on another scroll (PHerc. Paris 4).
- Still unknown: The text is fragmentary because only the inner core survives; systematic recovery across the library is described as future work.
- AI role: Machine learning used for ink detection on synchrotron scans, alongside improved computational unrolling.
Evidence sources: https://arxiv.org/abs/2606.29085
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: other. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 4/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The Nature paper, published online on 2026-06-24, reports a deep-learning ECG biomarker evaluated retrospectively in Swedish data and external US and Taiwanese datasets. Its high-risk group comprised 2.2% of the sample and had a 7.0% annual sudden-cardiac-death rate, versus 4.6% for reduced LVEF; 86.1% were not flagged by LVEF. Among high-risk patients with defibrillators, sudden cardiac death was 54.4% lower than model-expected, an observational association subject to confounding and not proof of benefit.
Technology: Deep learning ECG model with a generative ECG waveform model for visualization
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Existing Swedish ECG/death-certificate data, a held-out lockbox created before analysis, external US ventricular-arrhythmia and Taiwanese arrhythmic-arrest evaluations; observational defibrillator associations.
- Still unknown: Main paper reviewed; supplementary attachments not read. The authors explicitly identify residual confounding and the need for randomized testing of defibrillator benefit in model-high-risk patients.
- AI role: Deep learning model discovered the predictive ECG waveform pattern; generative model used to visualize it.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.1038%2Fs41586-026-10674-6%22&resultType=core&format=json · https://www.ebi.ac.uk/europepmc/webservices/rest/PMC13323061/fullTextXML
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 2/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A screen of about six million compounds found candidates active against resistant Neisseria gonorrhoeae. A1 reduced bacteria in animal models, and proteomics, genetics, and biochemistry confirmed its target, alanine racemase. These are preclinical results, published in Science Translational Medicine.
Technology: Deep learning for molecule screening
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Phenotypic screen of 38,650 molecules trained a graph neural network, which screened about 6 million compounds; 213 were validated experimentally, 83 inhibited growth. Two hits were tested in vitro against multidrug-resistant strains, by proteomics (alanine racemase target), in a vagina-on-a-chip model and in a mouse vaginal infection model.
- Still unknown: Not reported (the abstract describes the in vivo and organ-chip results as showing only early promise).
- AI role: A graph neural network trained on phenotypic screening data, used to virtually screen the compound library before experimental validation.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1126/scitranslmed.ads4699&resultType=core&format=json · https://news.ki.se/ai-helps-identify-new-antibiotic-candidates-against-drug-resistant-gonorrhoea
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
After experimentally screening activity-enhancing single mutants of LdTPS5, the researchers used PIFS-PLM, a few-shot model that selects features from protein-language-model embeddings using active-site and conserved-residue information, to propose mutation combinations. All 10 tested model-predicted combinations were active; nine produced more than 100 mg/L bicyclogermacrene in yeast, and the top variant produced 213.59 mg/L versus 2.93 mg/L for wild type (about 73-fold higher yield). A related selectivity-optimized combination increased bicyclogermacrene selectivity by 15%.
Technology: PIFS-PLM (Physics-Inspired Feature Selection of Protein Language Models), using ESM-2 embeddings with a few-shot mutation-combination classifier
Family ESM / ESMFold: profile · models page
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: After experimentally screening activity-enhancing single mutants of LdTPS5, the researchers used PIFS-PLM, a few-shot model that selects features from protein-language-model embeddings using active-site and conserved-residue information, to propose mutation combinations.
- Still unknown: The 73-fold result is product yield in the reported yeast system, reflecting effects of enzyme activity and expression; it is not a 73-fold increase in catalytic rate. The selectivity improvement was 15%. The model relies on a defined active site/conserved-site set and a small, enzyme-specific experimental dataset. Product profiles differed between cell-free extracts and yeast; these results do not establish universal performance across enzymes or production settings.
- AI role: After single substitutions were screened and activity-enhancing mutations experimentally validated, PIFS-PLM used 60–100 measured combinations to learn from selected ESM-2 features at active-site and conserved residues and predict beneficial combinations. Ten predicted LdTPS5 variants were tested; all were active, and the top variants substantially increased measured yeast product yield. The model also guided a selectivity optimization test. Initial single-mutation nomination used MM/GBSA or sequence conservation analysis, separate from PIFS-PLM combinatorial predictions.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13502649/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A neural network trained on observations from five Arctic field campaigns reduced ERA5 errors in the Arctic sea-ice surface energy budget and removed ERA5’s wintertime near-surface warm bias of about 4 K. The corrected estimates imply about 50% weaker winter surface cooling and 80–120 cm seasonal thermodynamic ice growth, closer to SMOS–CryoSat thickness increases than ERA5’s 150–200 cm estimate.
Technology: Feed-forward neural network product SEBai
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: A neural network trained on observations from five Arctic field campaigns reduced ERA5 errors in the Arctic sea-ice surface energy budget and removed ERA5’s wintertime near-surface warm bias of about 4 K.
- Still unknown: Validation uses leave-one-campaign-out cross-validation and multiple field campaigns, but direct flux observations remain sparse in space and time. The climatology applies to ice-covered Arctic regions north of 70°N with sea-ice concentration above 80%; authors do not recommend extrapolation to marginal ice zones or coastal regions. The ice-growth values are thermodynamic estimates rather than direct pan-Arctic measurements.
- AI role: The network learned the relationship between ERA5 atmospheric states and observed fluxes from MOSAiC, SHEBA, AO2018, N-ICE2015 and ARTofMELT campaigns, then bias-corrected hourly Arctic sea-ice surface fluxes for 1994–2024 to derive a revised regional climatology and ice-growth estimates.
Evidence sources: https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2025GL121289
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The MutexaGPT multi-agent large language model system translated a cold-adaptation design request into molecular simulations and linker-variant proposals for bidomain amylase. The authors report experimentally validated variants with 1.7-fold and 3.7-fold activity enhancement at 0 °C. The preprint appeared on 19 June 2026; the peer-reviewed journal version followed on 10 September 2026.
Technology: Multi-agent large language model system coordinating high-throughput molecular dynamics simulations and mutation design
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: The primary article reports two MutexaGPT-guided bidomain amylase linker variants with experimentally measured 1.7-fold and 3.7-fold activity enhancement at 0 °C. The Methods define relative activity using the 0 °C to 45 °C activity ratio.
- Still unknown: The experimental result is specific to the tested bidomain amylase and variants; broader enzyme-family transfer and independent reproduction are not established.
- AI role: LLM agents clarify the design request, translate it into computational plans and mutation patterns, and explain simulation outputs. High-throughput molecular dynamics metrics guide the proposed variants; experimental assays test their activity.
Evidence sources: https://assets-eu.researchsquare.com/files/rs-10072540/v1_covered_65be6e89-30da-488a-a7d7-0310d3a07d5f.pdf · https://www.nature.com/articles/s43588-026-01049-y
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary Research Square v1 full-text indexed content and metadata; primary Nature Computational Science indexed article text, including abstract, Methods, experimental result, date and affiliation block..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
APEX 1.1 screened 19.3 million peptide fragments from 2,897 prion-related proteins and identified 1,179 candidate antimicrobial peptides, termed prionins. Among 75 synthesized candidates, 59 inhibited at least one tested bacterial strain; two selected peptides reduced Acinetobacter baumannii burden in a mouse skin-abscess model. The study shows that prion-related sequences can yield experimentally active antimicrobial fragments, but does not show that these peptides are naturally released or function as host-defense agents in their source organisms.
Technology: APEX 1.1 deep-learning antimicrobial-peptide prediction and prioritization
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Of 75 synthesized APEX-selected prionins, 59 inhibited at least one bacterial strain in vitro; two selected peptides reduced A. baumannii burden in a mouse skin-abscess model.
- Still unknown: No evidence shows these fragments are naturally released from their source proteins or physiologically function as host-defense peptides in native organisms. Mouse results do not establish human efficacy or safety.
- AI role: APEX 1.1 scored peptide fragments for predicted antimicrobial activity and prioritized candidates from a search across prion-related proteins for synthesis and experimental screening.
Evidence sources: https://www.nature.com/articles/s41564-026-02408-1
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary Nature Microbiology full text as indexed, including methods, results, author contributions and affiliations..
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Among 376 unresolved cases, o3 Deep Research proposed explanations reviewed by specialists: 18 diagnoses were clinically confirmed, seven of which had already been established elsewhere but were missing from local records. This was a retrospective study across heterogeneous cohorts, without a measure of time saved or effect on care; clinicians made the decisions.
Technology: o3 Deep Research
Family o1 / o3: profile · models page
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Retrospective multicohort reanalysis of 376 unresolved cases; a diagnosis required a pathogenic or likely pathogenic variant confirmed in a CLIA-certified lab and returned to families, and seven rediscoveries were also reported.
- Still unknown: The abstract states that the retrospective design motivates prospective multicenter evaluation with predefined endpoints, calibration reporting and comparator baselines.
- AI role: An LLM ingested clinician notes, HPO terms and a filtered variant table and proposed candidate hypotheses for expert adjudication under ACMG/AMP criteria; experts made the classifications.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1056/AIcs2501343&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A DeepSequence-based predictor ranked SpuFz1 mutations without experimental activity labels; experimentally confirmed substitutions were combined into SpuFz1-v3. At the B2M locus in HEK293T cells, the article reports 35.52% indel formation versus 3.07% for wild type. Comparative sequence analysis, not this AI model, guided the separate ultrashort RNA design. The reported activity is assay- and locus-specific; off-target and clinical safety are not established.
Technology: DeepSequence-based Fanzor-Fitness Predictor
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The primary source reports the study-specific measured result summarized here.
- Still unknown: Independent replication and clinical or industrial efficacy are not established.
- AI role: A learned sequence fitness predictor ranks candidate mutations; researchers assay single substitutions and combine successful variants.
Evidence sources: https://www.nature.com/articles/s41467-026-74624-6
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Primary source indexed full text or abstract and publication metadata reviewed; scope stated in source_checks..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The system identified TEMPO as a useful additive for coupling primary sulfonamides; chemists reproduced yield improvements in 11 of 14 substrate pairs at bench scale. The work retains human supervision and specialized infrastructure; independent reproducibility and generalization remain to be checked.
Technology: GPT-5.4 and Maria
Family GPT-5: profile · models page
Publication: Announcement only · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: direct test
- Tested: Over two cycles in an automated lab, yields improved for 88% of boronic acids and 83% of sulfonamides tested (mean 16.6% to 25.2%); human chemists repeated representative reactions at bench scale, with higher yields for 11 of 14 substrate pairs.
- Still unknown: Not stated in the reviewed material
- AI role: GPT-5.4 connected to Molecule.one Maria proposed, designed and analyzed experiments; humans steered prompts, selected proposals, made limited corrections and validated the final result.
Evidence sources: https://openai.com/index/ai-chemist-improves-reaction/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A machine-learning pre-screen of candidate compounds, followed by first-principles calculations, selected YRu3B2 and LuRu3B2, which were measured as bulk superconductors with Tc of 0.81 K and 0.95 K and near 100% superconducting volume fraction. Evidence caveat: these are low-temperature superconductors and the work is a proof-of-concept screening workflow, not a room-temperature result; LuRu3B2 was initially overlooked because of a phonon instability.
Technology: Machine-learning pre-screening combined with first-principles calculations
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The Rice team synthesized YRu3B2 and LuRu3B2 and tested them to confirm superconductivity; the page gives no transition temperatures.
- Still unknown: A proof-of-concept study; the page does not report operating temperature, and most viable-looking combinations are described as unusable.
- AI role: Machine learning narrowed elemental combinations, then a pre-screening algorithm and targeted calculations selected candidates.
Evidence sources: https://www.aalto.fi/en/news/researchers-identify-new-superconductors-unlocking-process-that-could-yield-thousands-more
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A LightGBM classifier trained on routine clinical and laboratory data distinguished serum immunofixation-positive monoclonal immunoglobulin cases from negative cases. In an independent three-hospital cohort of 2,270 patients referred for immunofixation testing, the model achieved an AUC of 0.945, with 87.0% sensitivity and 89.6% specificity. In a subset with serum protein electrophoresis results, model sensitivity was 84.6% versus 75.3% for electrophoresis; in a separate 627-patient time-extended cohort, sensitivity was 88.2%, above each of three nonspecialist clinicians in that study.
Technology: LightGBM classifier selected from eight machine-learning models
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: In an independent three-hospital cohort of 2,270 patients referred for immunofixation testing, the model achieved an AUC of 0.945, with 87.0% sensitivity and 89.6% specificity.
- Still unknown: This is retrospective diagnostic validation, not evidence that screening with the model improves health outcomes or shortens diagnosis in practice. All participants had been referred for immunofixation testing at tertiary or regional hospitals; an unselected community or primary-care population was not studied. The 627-patient later cohort was time-extended but not strictly prospective. The model is an adjunct for flagging patients for confirmatory tests; immunofixation remained the reference standard, and the paper states broader population validation and clinical utility remain to be established.
- AI role: Researchers used five feature-selection algorithms to identify shared predictors and trained/evaluated eight classifiers on routine clinical and laboratory measurements. The selected LightGBM model produced patient-level MIg-positive probabilities; its diagnostic performance was measured against serum immunofixation results in independent multi-center and later-collected cohorts, including direct comparisons with serum protein electrophoresis and nonspecialist clinicians.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13314525/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Fung-AI combines a generative adversarial network with learned antifungal and hemolysis classifiers to select peptides for testing. The study first tested five peptides from separate clusters; activity of peptide 40 prompted testing of eight more members from cluster 17. Five of the nine tested cluster-17 peptides inhibited Fusarium graminearum, a post-selection cluster-specific outcome—not a general 5/13 model hit rate. Four peptides showed activity against Candida albicans, while none inhibited C. auris up to 500 μg/mL. The paper’s HepG2 cytotoxicity statements are internally inconsistent for peptide 48 and peptide 65, and peptide 12 and 40 have reported cytotoxicity; these assays establish neither human safety nor clinical efficacy.
Technology: Fung-AI GAN; learned antifungal and hemolysis classifiers
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Primary article reports staged testing, fungal MICs and cell-line cytotoxicity; adaptive denominator and conflicting LC50 statements were reconciled in the continuation audit.
- Still unknown: Sequence-level prior art and independent replication not assessed; HepG2 assays do not establish human safety.
- AI role: GAN generates candidates; antifungal and hemolysis classifiers narrow the candidate set before synthesis and fungal assays.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13289945/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 16 Nov 2026. Review scope: Primary PLOS article and indexed PMC full text checked; follow-up audit read full article and resolves staged denominator and LC50 inconsistency, not by remeasurement..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A multimodal literature-mining workflow and a physics-guided deep ensemble screened more than 150 million hypothetical compositions. The study reports that the x=0.01 and x=0.02 members of (Sr0.48Na0.26Bi0.26)(Ti1−xSnx)O3 met the authors’ predefined dielectric-stability targets while retaining room-temperature permittivity above 3300. The parent composition family and Sn substitution were already known; the result is a constrained low-Sn composition window, not a new parent chemistry. The reported upper-temperature stability criteria do not establish device certification or deployment.
Technology: Multimodal literature mining; deep neural-network ensemble
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The authors report experimental validation of the x=0.01 and 0.02 Sn compositions against predefined dielectric-stability targets and room-temperature permittivity above 3300.
- Still unknown: The parent family and substitution were known; the stated criteria do not establish full device certification or deployment. Direct publisher page was inaccessible, so evidence is limited to PubMed metadata and abstract in this check.
- AI role: Multimodal literature mining curated prior measurements and a deep neural-network ensemble screened candidate dielectric compositions and guided refinement.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/42297784/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 16 Nov 2026. Review scope: Primary PubMed record, article abstract, first-online date and affiliations..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The June 2026 bioRxiv version describes an initial machine-learning-guided screen of ZINC20 compounds balancing predicted antibacterial activity, mammalian-cell toxicity, and mechanism-of-action novelty. Among 61 purchased candidates, six showed partial growth inhibition; 49 were non-toxic to HepG2 cells at 50 μM. Flow cytometry indicated membrane disruption for four of the six initial active hits and membrane depolarization for two. A September version reports additional screening and revised models; its expanded results are not attributed to the June version here. The study does not report a novel protein target or clinical antibiotic.
Technology: Graph-convolutional neural networks for antibacterial activity/toxicity with XGBoost mechanism-of-action classification and novelty filtering
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: The June 2026 v1 reports that 61 purchased candidates were tested, six showed partial growth inhibition, and 49/61 were non-toxic to HepG2 cells at 50 μM. Flow cytometry indicated membrane integrity disruption for four initial active hits and membrane depolarization for two.
- Still unknown: The initial work does not report a novel protein target and is limited to the stated bacterial assays and HepG2 toxicity concentration. The September version adds expanded screening and revised analyses; those later results are not attributed to the June v1.
- AI role: Learned activity and toxicity models, together with a mechanism-novelty classifier, prioritized compounds for purchase and experimental testing.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13278111/ · https://pubmed.ncbi.nlm.nih.gov/42327090/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary PMC-indexed full text for the June bioRxiv v1, primary PubMed preprint abstract/metadata, and primary PMC/PubMed metadata for the later September version to distinguish version-specific results..
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
RFdiffusion-based de novo binder design and experimental screening produced synthetic proteins that inhibited three bacterial defense systems: Thoeris and Avs1 in Bacillus subtilis assays, and DdmDE in Vibrio cholerae. Five designed proteins inhibited Thoeris and nine inhibited Avs1 in the reported screens. An engineered SBSphiJ phage carrying both anti-Thoeris and anti-Avs1 proteins replicated in B. subtilis cells expressing both defenses. In native V. cholerae O1 El Tor, two of four tested DdmDE binders improved plasmid resistance to clearance.
Technology: RFdiffusion de novo binder design, ProteinMPNN sequence design and AlphaFold2/3 structure-guided candidate scoring
Family AlphaFold: profile · models page
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: RFdiffusion-based de novo binder design and experimental screening produced synthetic proteins that inhibited three bacterial defense systems: Thoeris and Avs1 in Bacillus subtilis assays, and DdmDE in Vibrio cholerae.
- Still unknown: These are laboratory proof-of-concept results in three bacterial defense systems and specific strains; they do not demonstrate therapeutic phage efficacy in animals or humans. Binder success was low relative to the computationally designed candidate sets, and the authors note that a verified binder may not work against homologous defense proteins. The native V. cholerae result was tested for plasmid retention, not general transformation across clinical isolates. The earlier bioRxiv version is the same study and is not a separate record.
- AI role: RFdiffusion generated candidate binder backbones for selected target sites in the Thoeris, Avs1 and DdmDE defense proteins; ProteinMPNN designed sequences and AlphaFold-based co-folding/scoring prioritized candidates for experiments. The selected synthetic binders were tested for defense inhibition, direct interactions and ability to restore phage infection or plasmid retention. The measured anti-defense functions and engineered-phage outcomes are the paper-specific results enabled by AI-guided protein design.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13402655/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Composition-based machine-learning models screened salt-hydration reactions, after which researchers synthesized and measured selected candidates by differential scanning calorimetry. The article reports volumetric energy densities of 1.74 ± 0.02, 1.73 ± 0.02 and 1.70 ± 0.02 GJ/m³ for three rubidium Tutton salt hydrates. These compounds were already known; the bounded result is AI-guided thermochemical characterization, not new compound synthesis or demonstrated heat-storage deployment. An earlier preprint contains the measurements, but its precise public date/version chronology is unresolved, so no first-public claim is made.
Technology: Stoichiometry-based machine-learning models of salt hydration properties
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Primary article reports the described measurements.
- Still unknown: Independent replication and global priority are not established.
- AI role: Stoichiometry-based machine-learning models of salt hydration properties
Evidence sources: https://www.nature.com/articles/s43246-026-01200-2
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Primary publisher full text and article metadata reviewed for stated result, AI contribution, date and affiliation..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
An expanded machine-learning earthquake catalog and complementary receiver-function and ambient-noise images reveal a previously unrecognized, roughly 250-km linear cluster of about 1,750 earthquakes that the authors interpret as tracing the northeastern/eastern edge of the subducted Yakutat microplate beneath south-central Alaska. The inferred slab edge lies beneath the curved Denali fault; its possible role in the 2002 Mw 7.9 earthquake’s nucleation is a hypothesis, not an established cause.
Technology: Modified EQTransformer machine-learning seismic phase picker, REAL phase associator, and NonLinLoc locator
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: An expanded machine-learning earthquake catalog and complementary receiver-function and ambient-noise images reveal a previously unrecognized, roughly 250-km linear cluster of about 1,750 earthquakes that the authors interpret as tracing the northeastern/eastern edge of the subducted Yakutat microplate beneath south-central Alaska.
- Still unknown: This is a localized tectonic interpretation from data collected between 2018 and 2021, not prospective earthquake prediction. The 64,870-event catalog reports average location uncertainties of about 5.0 km latitude, 5.2 km longitude and 8.6 km depth; the authors support the slab-edge interpretation with seismicity, receiver functions, tremor distribution and ambient-noise tomography. They propose—but do not establish—that the slab edge influenced nucleation of the 2002 Denali earthquake and the position of volcanic fields.
- AI role: The ML phase picker helped construct a denser earthquake catalog from permanent and temporary station records. The 64,870-event catalog exposed the previously unseen long linear earthquake cluster; receiver-function and ambient-noise tomography supplied independent complementary images used to interpret its relationship to the Yakutat slab edge.
Evidence sources: https://pubs.geoscienceworld.org/ssa/tsr/article/6/2/230/731510/Razor-Sharp-Edge-The-Yakutat-Slab-Dissecting-South
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
University of Chicago researchers trained a GPT model with a new notation (fLine) to generate complete battery electrolyte formulations, including solvent ratios, salt concentrations and temperature. Several AI-proposed compositions were synthesized and, per the release, performed on par with the state of the art in lithium metal batteries. No cycling or capacity figures are given, the work is early, and generated candidates still need experimental verification.
Technology: ElectrolyteGPT (GPT with fLine notation)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Generated solvents and formulations were synthesized and fabricated, found to meet the target property requirements, and enabled long-term cycling in anode-free lithium metal batteries.
- Still unknown: The release states the limited data and parameter set, and that the results matched but did not exceed top existing electrolytes; the abstract states no limitations.
- AI role: A transformer-based generative model (ElectrolyteGPT) generated candidate solvents and electrolyte formulations conditioned on property targets.
Evidence sources: https://news.uchicago.edu/story/electrolytegpt-can-generate-new-formulations-battery-development · https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1021/jacsau.5c01628&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Researchers screened 1,513 natural SGNH-hydrolase homologues and trained task-specific sequence CNNs on the resulting activity, thermostability, and substrate-preference data. The models helped select candidates for a prospective 230-sequence assay panel, where CNN-selected candidates were enriched for target traits relative to generalist pretrained and physics-based baselines; the panel also included other selection strategies and random controls. Assays used clarified E. coli lysates and measured relative maximum velocity, integrating expression and catalysis. The first result-bearing version was posted to bioRxiv on 2026-06-03; a Research Square version followed on 2026-06-05.
Technology: Task-specific convolutional neural networks trained on amino-acid sequences; CNN-based HEV and UCB selection strategies
Publication: Preprint · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: prospective or independent check
- Tested: A prospective panel of 230 previously untested sequences was assayed; CNN selection strategies were enriched for the predicted target traits compared with generalist pretrained and physics-based approaches. The panel pooled sequences selected by multiple strategies and included random controls.
- Still unknown: Clarified lysate MaxV integrates enzyme expression and catalytic activity and is not a purified-enzyme turnover measurement. The result is scoped to this enzyme family and the reported assay setup; prospective panel composition was mixed across selection strategies.
- AI role: Supervised family-specific CNNs trained on sequence and assay data ranked untested homologues for activity, thermostability, and substrate preference, and guided candidate selection for prospective experimental testing.
Evidence sources: https://www.biorxiv.org/content/10.64898/2026.06.02.729712v1.full
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary bioRxiv v1 full text as indexed, including title, posting date, affiliations, methods, prospective assay results and limitations; primary Research Square v1 identity/date record..
Editorial assessment: Impact 3/5 · Validation 4/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A supervised Random Tree classifier on satellite images predicted a drop of over 90% in live Acropora cover at Belize from 2011 to 2023 across 25 km². Field searches in June 2025 found live or recently dead colonies at 28 of 42 predicted sites and live colonies at 16 previously undocumented locations. The field check was a search for survivors at sites the model flagged, not a blind test.
Technology: Random Tree supervised classification of satellite imagery (machine learning)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: A Random Tree classifier's predicted over-90% decline in Acropora cover (2011-2023) was compared with field data from 2012-2025, and field searches at 42 predicted sites found live or recently dead colonies at 28 of them.
- Still unknown: The abstract states no limitations; the search was a targeted check of flagged sites rather than a blind test, and 14 of 42 predicted sites had no colonies found.
- AI role: A supervised Random Tree classifier identified Acropora in satellite images and generated the cover predictions that guided the field searches.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1038/s41598-026-47081-w&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The Crossbreeding Neural Network (CBNN), trained jointly on carbon-supported single-atom catalysts and perovskite oxides, screened 8,008 candidates. Its top W-Mo-Ru-Rh design on a perovskite support outperformed the comparison catalysts in electrochemical tests, and the activity ranking of 12 tested catalysts matched the predictions. These counts come from the IBS institutional release; the paper abstract confirms the experimental result but does not quantify the activity gain. The catalog date is the Nature Materials publication date, 28 May 2026.
Technology: CBNN (Crossbreeding Neural Network)
Publication: Announcement only · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: direct test
- Tested: The AI-predicted catalysts were synthesized and tested for alkaline oxygen evolution; the activity ranking of 12 tested catalysts matched the predictions.
- Still unknown: The comparison is limited to catalysts synthesized in this study; use in batteries, energy storage and drug discovery is not demonstrated.
- AI role: A Crossbreeding Neural Network trained on carbon-supported single-atom catalysts and perovskite oxides screened 8,008 multimetallic candidates.
Evidence sources: https://www.ibs.re.kr/cop/bbs/BBSMSTR_000000000738/selectBoardArticle.do?nttId=26725&pageIndex=1&searchCnd=&searchWrd=
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A deep-learning analysis of regional Antarctic seismic records identified more than 500 previously unrecognized intermediate-depth earthquakes roughly 100–150 km beneath David Glacier in East Antarctica. The earthquakes occur within a tectonic plate rather than at a plate boundary. The authors explain the concentration through bending stresses associated with variable thermally driven uplift and an abrupt change in lithospheric strength; the ice sheet may also affect regional stress.
Technology: Deep-learning seismic-event detector applied to regional seismic records
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A deep-learning analysis of regional Antarctic seismic records identified more than 500 previously unrecognized intermediate-depth earthquakes roughly 100–150 km beneath David Glacier in East Antarctica.
- Still unknown: The discovery concerns a newly recognized concentration of intraplate events, not a new earthquake mechanism conclusively proven. Proposed stress sources are interpretations; the paper notes similar events elsewhere may remain undetected.
- AI role: Detected intermediate-depth earthquake signals in regional seismic data that were not recognized in prior catalogs; the authors then located and interpreted the events using seismological and geodynamic analysis.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/42207895/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A composite score combines Boltz2-predicted binding and DLKcat-predicted turnover to rank variants in a focused Agr5 library. Of 15 top-ranked variants tested, four produced more viridiflorol than wild-type Agr5; the paper reports 11.05-fold higher production for Y374F and a 952.79 mg/L titer in its yeast assay. The initial V314G design involved mechanism-guided human analysis, and the measured values are assay-specific.
Technology: DeEnzyme_score combining Boltz2 and DLKcat predictions
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The article reports four activity-enhancing variants among 15 tested and specifies the Y374F production ratio and titer in a yeast assay.
- Still unknown: Prior sequence-level overlap, earlier disclosure and independent replication remain unestablished; the proposed allosteric explanation partly relies on simulation.
- AI role: Boltz2 and DLKcat predictions are combined to rank variants selected for experimental testing; the initial V314G design is attributed to mechanism-guided analysis, not AI.
Evidence sources: https://pubs.acs.org/accacs/article/16/14/13022/5114435/Customizing-Natural-Products-of-Sesquiterpene
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Publisher indexed full article text including title, methods, results, article history and author-affiliation block; direct ACS endpoint returned 403..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Spatial-DC reconstructed cell-type-resolved spatial proteomes and suggested COL8A1–ITGA2 communication between cancer-associated fibroblasts and pancreatic cancer cells. Multiplexed immunohistochemistry supported the spatial predictions; human protein co-expression was associated with survival. The pairing was already known: the result concerns its inferred spatial deployment, not a new receptor pair or established causal mechanism. This record uses the May 27, 2026 journal version; the result already appeared in the December 20, 2024 preprint.
Technology: Spatial-DC; graph-enhanced transfer learning
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Retrospective spatial proteomic reconstruction and human survival association. Same-study multiplexed immunohistochemistry supports spatial distributions and co-expression, not causal COL8A1–ITGA2 communication.
- Still unknown: No causal signaling perturbation, therapeutic benefit or independent replication is established in the reviewed material. Current review read indexed publisher introduction and NLM abstract/metadata; full journal methods were inaccessible. Earlier primary preprint XML was read in the project audit on November 16, 2026; revised journal quantitative details were not assessed.
- AI role: Graph-enhanced transfer learning deconvolved mixed spots and reconstructed cell-type-resolved protein profiles used to infer spatial interactions.
Evidence sources: https://www.sciencedirect.com/science/article/abs/pii/S2405471226000943 · https://pubmed.ncbi.nlm.nih.gov/42202779/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
An agent that alternates LLM proof generation with Lean verification autonomously resolved 9 of 353 open Erdős problems and proved 44 of 492 OEIS conjectures, at a cost of a few hundred dollars per problem according to the abstract. Proofs are Lean-verified, so the checking is mechanical; the abstract does not name the system and the results are limited to the problem sets tested.
Technology: LLM-based agent with Lean-based proof verification (system unnamed in abstract)
Publication: Preprint · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: prospective or independent check
- Tested: Proofs generated by the agent were checked in Lean; the agent resolved 9 of 353 open Erdos problems and proved 44 of 492 OEIS conjectures, at a stated cost of a few hundred dollars per problem.
- Still unknown: The abstract does not describe how the problem sets were selected, and notes the simpler agent cost more on the hardest problems.
- AI role: An agent alternates LLM proof generation with Lean verification to prove open problems autonomously.
Evidence sources: https://arxiv.org/abs/2605.22763
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 4/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A theory-guided inverse-design study combined a crystal diffusion variational autoencoder with a Crystal Transformer to propose a body-centered tetragonal C24 allotrope, Netsene. First-principles calculations report dynamical and mechanical stability and a nested nodal-surface semimetal band structure with drumhead surface states. This is a computationally predicted allotrope and electronic phase, not an experimentally synthesized material.
Technology: Crystal diffusion variational autoencoder (CDVAE) and Crystal Transformer, guided by symmetry-engineering theory
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: A theory-guided inverse-design study combined a crystal diffusion variational autoencoder with a Crystal Transformer to propose a body-centered tetragonal C24 allotrope, Netsene.
- Still unknown: The allotrope and its properties are computational predictions validated by first-principles calculations; the paper does not report synthesis or experimental observation. The finding is one proposed material/phase result from one paper, not a claim that AI alone established the physical existence of Netsene.
- AI role: Generated and ranked candidate crystal structures within a theory-guided inverse-design workflow; the resulting Netsene structure was then assessed with first-principles calculations for stability and topological band properties.
Evidence sources: https://arxiv.org/abs/2605.22141
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A random-forest model using six terrain morphometrics from ArcticDEM classified glacial erosional styles across Arctic Canada and Greenland. The map indicates that regions with little or no glacial erosion, and thus potentially preserved pre-glacial landscape fragments, occur more widely than previous mapping recognized; it also finds distinct morphologies between Greenland and Canada. The authors infer that inherited topography, ice dynamics, latitude and geology help shape these patterns.
Technology: Random-forest machine-learning classifier
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A random-forest model using six terrain morphometrics from ArcticDEM classified glacial erosional styles across Arctic Canada and Greenland.
- Still unknown: The paper reports validation accuracy of about 79.5% (±4.6%) and notes low classification probabilities in some areas, where training coverage or coarse classes may limit confidence. “Little or no glacial erosion” is a model classification interpreted as possible preservation of pre-glacial terrain, not direct dating or field confirmation of every mapped tile. Proposed controls such as ice dynamics and basal hydrology are interpretations, not demonstrated causal mechanisms.
- AI role: The random forest classified ArcticDEM tiles using six elevation-derived morphometric variables, assigning class probabilities at continental scale and producing model-derived feature-importance estimates. This enabled a quantitative revision of prior qualitative landscape maps and the reported finding that minimally glacially modified terrain is more widespread than previously mapped.
Evidence sources: https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2025JF008661
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
OpenAI presents a construction of planar point configurations with more unit-distance pairs than predicted by the Erdős conjecture. The announcement says mathematicians outside the team checked it and provides a complementary explanation; this record checks the source, not an independent reconstruction of the proof.
Technology: OpenAI internal model
Publication: Announcement only · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: prospective or independent check
- Tested: The proof was checked by a group of external mathematicians, who also wrote a companion paper explaining the argument.
- Still unknown: Not stated in the reviewed material
- AI role: A general-purpose internal OpenAI reasoning model produced the proof while being evaluated on Erdős problems.
Evidence sources: https://openai.com/index/model-disproves-discrete-geometry-conjecture/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 5/5 · Validation 4/5 · Novelty 5/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The IRB Barcelona team trained cell-line cytotoxicity predictors on data from more than 11,000 compounds assayed across pancreatic-cancer and control cell lines, then integrated the models with a reinforcement-learning generative system to design molecules for cell-selective effects. In follow-up tests, 45 AI-designed molecules or close analogues were assayed; 20 showed the intended cell-specific cytotoxicity, and dose-response testing confirmed 18 of those 20. Across the seven selectivity exercises, at least one hit was found in five; measured IC50 values could be calculated for 12 confirmed hits and ranged from 0.97 to 16.6 μM in the targeted cell lines.
Technology: Cell-line-specific cytotoxicity prediction models coupled to a reinforcement-learning molecular generator (REINVENT-based workflow)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The IRB Barcelona team trained cell-line cytotoxicity predictors on data from more than 11,000 compounds assayed across pancreatic-cancer and control cell lines, then integrated the models with a reinforcement-learning generative system to design molecules for cell-selective effects.
- Still unknown: This is a proof-of-concept hit-discovery study using in-vitro cell-line assays, not evidence of tumor selectivity in patients or an animal efficacy result. Only binary cytotoxicity labels at 10 μM informed the predictors; 45 generated structures or close analogues were tested, with 20 showing desired phenotypes and 18 confirming in dose-response assays. Exact generated structures were not always the synthesized compounds, and measured IC50 spans 0.97–16.6 μM for 12 of the confirmed hits. The paper explicitly characterizes this as an early hit-discovery stage.
- AI role: The authors trained predictive models on phenotypic screening data from >11,000 compounds, then used their scores as rewards/objectives in a generative workflow to propose molecules intended to kill one pancreatic-cancer cell line while sparing another or control cells. Candidate molecules or close purchasable analogues were experimentally tested and dose-response validated.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/42162204/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
In Gary Peltz's Stanford lab, two of the three existing drugs proposed by Co-Scientist blocked fibrosis and promoted regeneration in human liver cells, while Peltz's own two choices showed no benefit. The anticancer drug vorinostat blocked 91% of a damage response that can lead to scarring. Cell experiments only; published in Advanced Science.
Technology: Co-Scientist
Family AI Co-Scientist: profile · models page
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The anti-fibrotic efficacy and toxicity of 25 drugs were assessed serially in human liver organoids grown in microwells; Co-Scientist's candidates were tested in cells alongside the authors' own picks.
- Still unknown: Not reported.
- AI role: The AI co-scientist multi-agent system proposed three candidate drugs with reasoning, and helped direct their experimental characterization.
Evidence sources: https://deepmind.google/blog/uncovering-repurposed-medicines-to-fight-liver-fibrosis/ · https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1002/advs.202508751&resultType=core&format=json
The legacy label “Not reported” does not distinguish information absent from reviewed material from information not assessed. It is not a scientific finding of absence.
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Citations: in the top 10% of works of the same type, year and subfield (OpenAlex work metrics, 27 citations; as of 15 Nov 2026). Attention, not a measure of correctness; not part of the score.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Calico Life Sciences used Co-Scientist to formulate the hypothesis that metabolism regulates the cellular integrated stress response, refined experiments with the tool, and reports new findings. The post gives no numerical data, and the results have not yet been published or peer reviewed.
Technology: Co-Scientist
Family AI Co-Scientist: profile · models page
Publication: Announcement only · Test context: Test context unspecified in annotation. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: not reported
- Tested: Calico reports that experiments were run with the Co-Scientist hypothesis and produced new findings; the post does not describe the experiments, assays or results.
- Still unknown: Not reported (the results are not yet published; the team plans to publish them).
- AI role: Co-Scientist generated the hypothesis that metabolism regulates the integrated stress response, and was used to refine experimental design.
Evidence sources: https://deepmind.google/blog/opening-new-paths-in-aging-research/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 1/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Filippo Menolascina (Edinburgh) used Co-Scientist to search the literature for combinations of MASH therapies. The tool proposed the NLRP3 inflammasome as a link between inflammation and metabolism, and the post says the hypothesis was later tested experimentally. No numbers are given; the drug combinations were suggested for testing, not demonstrated.
Technology: Co-Scientist
Family AI Co-Scientist: profile · models page
Publication: Announcement only · Test context: Test context unspecified in annotation. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: not reported
- Tested: The article says the NLRP3 hypothesis was later experimentally verified but does not describe the experiments, who carried them out, or their results.
- Still unknown: Not reported (the article does not state limitations or whether the work is published).
- AI role: Co-Scientist searched the literature, proposed drug combinations for MASH therapy and proposed the NLRP3 inflammasome as the link between inflammation and metabolism.
Evidence sources: https://deepmind.google/blog/accelerating-discovery-of-liver-disease-mechanisms/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 1/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Omar Abudayyeh and Jonathan Gootenberg used Co-Scientist to search tens of thousands of papers, obtaining over 20 proposed genetic factors for reversing cellular aging, a couple of which were validated in the lab, and reducing analysis of genetic-screening data from six months to a few days. The post gives no sample size or peer review status.
Technology: Co-Scientist
Family AI Co-Scientist: profile · models page
Publication: Announcement only · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: direct test
- Tested: Lab tests validated a couple of the more than 20 proposed genetic factors, which pushed cells toward a younger state; the experiments and exact number are not described.
- Still unknown: Not reported (the post gives no sample size or peer-review status and contains only the investigators' quotes).
- AI role: Co-Scientist scanned tens of thousands of papers, proposed genetic factors for reversing cellular aging and analysed screening data alongside the literature.
Evidence sources: https://deepmind.google/blog/fast-tracking-genetic-leads-to-reverse-cellular-aging/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
WeatherNext forecast a Category 5 landfall in Jamaica five days ahead with 80% probability, rising to nearly 100% at three days. According to the post, it was the best single model for track and intensity in the National Hurricane Center's 2025 verification report, which used it alongside its HAFS model. The post gives no numerical comparisons; rapid intensification remains very hard to predict.
Technology: WeatherNext
Family WeatherNext: profile · models page
Publication: Announcement only · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: prospective or independent check
- Tested: The National Hurricane Center's 2025 annual verification report found WeatherNext the top-performing individual model for track and intensity; the Melissa forecast is presented as a single case.
- Still unknown: The page notes that predicting rapid intensification is very difficult and that Weather Lab is experimental; no dedicated limitations section.
- AI role: WeatherNext, an AI weather model with 50-member ensembles, forecast the Hurricane Melissa landfall alongside physics-based HAFS and was used by NHC forecasters.
Evidence sources: https://deepmind.google/blog/how-weathernext-helped-the-national-hurricane-center-better-predict-hurricane-melissas-historic-landfall-in-jamaica/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 4/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Google's Empirical Research Assistance (ERA), a Gemini-based tree-search system, reaches expert-level performance on six benchmarks spanning genomics, public health, satellite imagery, neuroscience, time-series forecasting and mathematics. Its state-level U.S. hospital admission forecasts for flu, COVID-19 and RSV rank at or near the top of CDC leaderboards. Results are computational and checked against public benchmarks and leaderboards, with no wet-lab validation; exact benchmark scores were not visible in the sources read.
Technology: Empirical Research Assistance (ERA), Gemini-based tree search
Family Gemini: profile · models page
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: ERA-generated methods and models were compared against human-developed methods on a public leaderboard for bioinformatics and against the CDC ensemble and other individual models for COVID-19 hospitalization forecasts.
- Still unknown: Not reported.
- AI role: A Gemini-based large language model with tree search generated scientific software, including 40 single-cell data analysis methods and 14 forecasting models.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1038/s41586-026-10658-6&resultType=core&format=json
The legacy label “Not reported” does not distinguish information absent from reviewed material from information not assessed. It is not a scientific finding of absence.
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 2/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
MIDAS, a multimodal graph-neural-network model, ranked candidate immuno-oncology targets from gene-interaction, patient multi-omic and immune-biology data. In eight stage III/IV melanoma patient-derived explants, researchers blocked OSM–OSMR signalling with an anti-OSM antibody. The paper reports reduced dysfunctional CD8+ T-cell populations and lower CCL4 levels; PTPN22 inhibition did not produce comparable changes in this assay.
Technology: MIDAS, a multimodal graph neural network for immuno-oncology target ranking
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The paper reports reduced dysfunctional CD8+ T-cell populations and lower CCL4 levels; PTPN22 inhibition did not produce comparable changes in this assay.
- Still unknown: Functional follow-up used only eight stage III/IV melanoma patient-derived explants. This is an ex-vivo preclinical result, not treatment of patients or evidence of clinical benefit. OSM–OSMR biology has prior literature in other tumour settings; the new paper-specific contribution is MIDAS target nomination and measured perturbation effects in melanoma explants. Do not generalize to other cancers or claim a validated therapy.
- AI role: MIDAS integrated gene-interaction networks with patient multi-omics, immune-cell biology and genetic-perturbation evidence to rank candidate immunotherapy targets. Researchers selected OSM–OSMR for functional follow-up and tested pathway blockade in patient-derived tumour explants; the model nominated the target, while the ex-vivo experiments measured the T-cell and cytokine outcomes.
Evidence sources: https://www.nature.com/articles/s42256-026-01201-3
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Follow-up spectra for 255 Gaia white dwarfs were compared with machine-learning classifications from low-resolution Gaia spectra. The study reports that only 5 of 112 objects classified as massive DBs were genuine DB white dwarfs; most were magnetic white dwarfs or warm DQ stars. Its spectroscopy also catalogued 63 magnetic white dwarfs and 29 warm DQs in the selected sample, while confirming high classification accuracy for major spectral types within the classifier’s training set.
Technology: Random-forest spectral classifier applied to Gaia low-resolution spectra
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: Follow-up spectra for 255 Gaia white dwarfs were compared with machine-learning classifications from low-resolution Gaia spectra.
- Still unknown: The result is a follow-up characterization of a selected 500-pc sample, not a complete census of white dwarfs. It corrects a specific machine-learning-labelled ‘massive DB’ subpopulation; reported counts should not be extrapolated to Gaia’s full white-dwarf population.
- AI role: The classifier assigned initial white-dwarf spectral types and selected objects of interest; astronomers obtained Gran Telescopio Canarias medium-resolution spectra and visually inspected them to determine the objects’ types and magnetic features.
Evidence sources: https://arxiv.org/abs/2605.16493
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Unsupervised clustering of 20 days of seismic data before the 28 May 2025 Blatten glacier collapse separated rockfall signals from events associated with basal glacier sliding. Sliding-related event counts accelerated approximately exponentially beginning about five days before collapse, with further acceleration about 1.5 days beforehand; the authors interpret this pattern as consistent with nucleation-like slip weakening.
Technology: Deep scattering network, independent component analysis and hierarchical clustering (unsupervised machine learning)
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: Unsupervised clustering of 20 days of seismic data before the 28 May 2025 Blatten glacier collapse separated rockfall signals from events associated with basal glacier sliding.
- Still unknown: The finding is a retrospective analysis of one event, relying primarily on a single seismic station 5.3 km from the site. The authors describe the physical mechanism as consistent with nucleation and hypothesize that rockfalls contributed to instability; these interpretations require study of other events. The inferred slip-rate proxy has high absolute uncertainty, and the approach is not an operational early-warning system.
- AI role: The unsupervised representation and clustering explored continuous single-station seismic records without event templates, separating rockfall, glacier-sliding and cultural-noise signatures. This made the evolving sliding-related event population and its acceleration measurable in the 20-day record.
Evidence sources: https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2025GL121175
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A regression model trained on experimental and theoretical data guided Mo/Sb/Br co-doping for a lead-free near-infrared phosphor. The authors report synthesizing Cs₂Zr(Cl₀.₄₆Br₀.₅₄)₆:12%Mo⁴⁺/3.6%Sb³⁺, with emission near 920 nm, internal quantum efficiency of 92.4%, and external quantum efficiency of 65.9% under the reported conditions. Its measured intensity was about 20% higher than the authors’ orthogonal-design comparator. A separate diffusion model was used for image reconstruction demonstrations; this does not establish clinical diagnostic performance. Earlier composition-level prior art and independent replication remain open.
Technology: Transformer-based regression model
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Authors report synthesis and optical characterization of the predicted composition, LED device testing, and separate imaging demonstration.
- Still unknown: Study-specific efficiency/intensity results; earlier independent reports and composition-level priority remain unverified; imaging is not clinical diagnostic validation.
- AI role: A model trained on experimental/theoretical data predicted co-dopant concentrations for researchers to synthesize and test; a separate diffusion model reconstructed imaging outputs.
Evidence sources: https://www.nature.com/articles/s41467-026-73105-0 · https://pubmed.ncbi.nlm.nih.gov/42140992/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 16 Nov 2026. Review scope: Primary indexed publisher article and NLM publication metadata; abstract/results, model roles, online date and measured outcomes..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Researchers trained cell-survival predictors on image-based screening data from six patient-derived glioblastoma stem-cell lines and screened more than 12,000 compounds. Experimental follow-up identified three compounds with activity across the GBM phenotype; HSP90 inhibitor XL-888 eliminated all six tested lines at nanomolar IC50 values, with analyses consistent with HSP90 inhibition and downstream client-protein disruption.
Technology: Machine-learning cell-survival predictors trained on high-throughput phenotypic screening
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Experimental follow-up identified three compounds with activity across the GBM phenotype; HSP90 inhibitor XL-888 eliminated all six tested lines at nanomolar IC50 values, with analyses consistent with HSP90 inhibition and downstream client-protein disruption.
- Still unknown: This is an in-vitro lead-discovery study across six patient-derived cell lines, not clinical efficacy. XL-888 is an existing HSP90 inhibitor and its proposed development as a GBM lead needs medicinal-chemistry optimization, including improved blood–brain-barrier penetration; no patient outcomes are reported.
- AI role: ML models trained on in-house imaging/compound-response data predicted cell survival and prioritized compounds from a >12,000-compound screen; experimental assays across patient-derived GBM stem-cell models validated three leads and tested mechanisms. Researchers performed the wet-lab work and mechanism analyses.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/42158926/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
BioGDR combines molecular structure with learned, predicted gene-expression and kinase-effect features and transcriptomic cell representations. Prospective in-vitro tests confirmed differential sensitivity to Golgicide A-2, including T84 colorectal cells (reported IC50 1.09 μM). This is a cell-line sensitivity result, not evidence of clinical benefit or discovery of the compound itself.
Technology: BioGDR multimodal interpretable deep learning; pathway graph attention
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The primary source reports the study-specific result summarized here.
- Still unknown: Independent replication and clinical benefit are not established by this source review.
- AI role: The model integrates predicted compound effects and cell transcriptomes to rank drug responses; the study prospectively tests candidate sensitivity in cell lines.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13396495/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Primary source text reviewed; source_checks specify accessible sections..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A candidate identified with MatterSim-v1, tetragonal tantalum phosphide, was synthesized and showed thermal conductivity of 152 W/m/K in the best samples, close to silicon. MatterSim-v1 inference is 3–5 times faster, and the new multi-task model MatterSim-MT predicts energies, forces, and properties. Battery and other property results are predictions, not experimental confirmations.
Technology: MatterSim-v1 and MatterSim-MT
Family MatterSim: profile · models page
Publication: Announcement only · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: direct test
- Tested: A MatterSim-v1 screen of over 240,000 candidates picked tetragonal tantalum phosphide, which was synthesized at UT Dallas and whose thermal conductivity was measured at UIUC, reaching 152 W/m/K in the best samples; the page does not describe the measurement method or number of samples.
- Still unknown: The page states that practical constraints such as mass density, elemental abundance and environmental stability are needed, but does not say whether TaP meets them.
- AI role: MatterSim-v1 machine-learning interatomic potentials estimated phonon thermal conductivity to screen candidates before experimental validation; MatterSim-MT is a multi-task model.
Evidence sources: https://www.microsoft.com/en-us/research/blog/advancing-ai-for-materials-with-mattersim-experimental-synthesis-faster-simulation-and-multi-task-models/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
VIBANN used a dataset of 673 bulk metallic-glass compositions to guide the design of five alloys, synthesized as amorphous 2 mm rods. The paper's experimental results identify alloy A1, B68Nb24Fe4W4, as the hardest: 2,447 +/- 44 HV at a 0.5 N indentation load, with six replicate indentations. Hardness decreases with increasing load; the peak is not a value for every alloy or test condition. X-ray scattering and molecular dynamics support a structural interpretation within the studied composition range. The date refers to the May 2026 paper, rather than the institute's later August announcement.
Technology: VIBANN (variational information-bottleneck attention network)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Designed bulk metallic glasses reached a Vickers hardness of about 2450 HV, compared with other alloys under comparable conditions.
- Still unknown: Discovery is constrained by the vastness of composition space and scarce validated data, and results apply under limited-data conditions.
- AI role: An uncertainty-aware latent-representation framework searched compositions, screened by chemical plausibility, novelty and predictive uncertainty.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1038/s41467-026-73008-0&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
In vitro metabolomics of 667 purified E. coli enzymes, analyzed with a machine-learning-derived multivariate classifier, indicated previously unrecognized promiscuous activities in 11% of enzymes tested. Follow-up reconstruction assigned 22 reaction stoichiometries to 11 enzymes, with four reactions experimentally validated; examples include expanded nucleotide-substrate use by DeoA, cofactor and sugar promiscuity in NanK, and FMN dephosphorylation by CobC. Follow-up assays also found that pyrimidine-specific ribonucleoside hydrolase RihB acted on thymidine, guanosine and deoxyuridine, further extending its measured substrate range.
Technology: Machine-learning-derived multivariate classifier applied to enzyme metabolomics ion-trace data
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: In vitro metabolomics of 667 purified E. coli enzymes, analyzed with a machine-learning-derived multivariate classifier, indicated previously unrecognized promiscuous activities in 11% of enzymes tested.
- Still unknown: The classifier operated at a 33% false-discovery rate; the 135 ion traces were putatively annotated, and only four of 22 reconstructed reaction stoichiometries were experimentally validated. Therefore the full 11% result is a set of candidate activities, not 11% independently confirmed enzyme functions. The RihB activities were prompted by a low-confidence potassium-thymidine-adduct trace and then tested directly with purified enzyme; they should be distinguished from the classifier’s higher-confidence hit set.
- AI role: A machine-learning-derived multivariate classifier, with a reported false-discovery rate of 33%, detected unexpected changes in metabolite ion traces in the enzyme assays and helped nominate candidate promiscuous activities. The researchers then reconstructed candidate reaction stoichiometries and experimentally validated four reactions.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13389410/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The MISPOP ensemble screened a natural peptide library of 1,033 sequences, prioritized 16 oncolytic-peptide candidates and selected five for synthesis and testing. Dermaseptin-S9 showed membrane disruption and immunogenic-cell-death readouts in vitro and over 92% tumor-growth inhibition in a B16F10 melanoma mouse model.
Technology: MISPOP mechanism-informed ensemble combining XGBoost, deep neural networks and transfer learning with physicochemical priors
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The MISPOP ensemble screened a natural peptide library of 1,033 sequences, prioritized 16 oncolytic-peptide candidates and selected five for synthesis and testing.
- Still unknown: The active peptide is a natural sequence rather than a new AI-designed molecule. Findings are preclinical: three tumor cell lines and a B16F10 mouse melanoma model (n=5 per group shown in article figures); they do not establish human efficacy or absence of toxicity beyond the tested model/doses. Only five of 16 model-prioritized candidates were synthesized and evaluated.
- AI role: The machine-learning ensemble integrated three predictors and mechanism-related physicochemical features to prioritize candidate oncolytic peptides from a natural sequence library. Five shortlisted peptides were synthesized and evaluated; Dermaseptin-S9 was tested in cell-based membrane-disruption/ICD assays and a mouse melanoma model.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13336043/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A machine-learning workflow catalogued 7,346 microearthquakes (ML −2.0 to 2.18) recorded during 2021 stimulation and circulation tests at China’s Gonghe enhanced-geothermal-system site. The catalog resolves a ring-shaped distribution with four clusters and varying faulting styles. The authors infer that site-scale seismicity reactivated mapped, pre-existing natural faults; at borehole scale, the principal seismogenic zone coincides with strong SHmax rotation. Stress modeling estimates a low fault-friction coefficient near 0.23 (95% CI 0.19–0.48) and the low-to-moderate fluid overpressure needed for slip.
Technology: Machine-learning earthquake-detection and cataloging workflow (model details are not stated in the publisher abstract)
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A machine-learning workflow catalogued 7,346 microearthquakes (ML −2.0 to 2.18) recorded during 2021 stimulation and circulation tests at China’s Gonghe enhanced-geothermal-system site.
- Still unknown: The source checked here is the original publisher abstract and plain-language summary. The catalog is observed, while fault reactivation, chlorite-rich gouge consistency, stress controls and required fluid overpressure are the authors’ interpretations/model estimates, not direct causal proof. No claim is made that the AI model alone establishes these mechanisms.
- AI role: The authors state that machine learning was used to detect earthquakes and compile the high-precision 7,346-event catalog; the resolved spatial pattern supplies the seismicity observations used in their scale-dependent fault-structure interpretation and stress analysis.
Evidence sources: https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2025JB033158
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A PhaseNet-assisted workflow combined phase detection, polarisation analysis and template matching on records from one permanent station to reconstruct seismicity in the two years before the 2025 Mw 7.1 Dingri earthquake. The authors report increased event rate and clustering beginning about 45 days before the mainshock. This retrospective pattern is interpreted as compatible with late-stage nucleation; the study does not establish a prospective earthquake forecast.
Technology: PhaseNet deep neural-network seismic phase picker combined with polarisation analysis and template matching
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: PhaseNet-assisted phase detection, polarisation analysis and template matching reconstructed 289 template events and 1,138 augmented events at one station before the 2025 Dingri earthquake. The authors report increased event rate and clustering beginning about 45 days before the mainshock.
- Still unknown: This is a retrospective analysis from one station. The authors describe a pattern compatible with late-stage nucleation; it does not establish a prospective forecast or earthquake prediction.
- AI role: PhaseNet, a deep neural-network phase picker, supported detection and event-catalogue construction; the resulting catalogue enabled retrospective identification of the pre-mainshock activity increase.
Evidence sources: https://www.nature.com/articles/s41598-026-49197-5
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary Scientific Reports indexed article, including abstract, Methods, Results, article dates and author-affiliation block..
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A closed-loop microfluidic platform ran 120 experiments in one 12-hour campaign on double perovskite nanoplatelets with up to six elements and identified the brightest lead-free composition found in that campaign. The press release gives no numerical photoluminescence gain, so the improvement is not quantified in the sources checked. Validation is the authors' own experimental measurements, and the paper should be checked for the brightness figures.
Technology: PoLARIS (microfluidic self-driving lab with closed-loop experiment selection)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Lab experiments in one 12-hour campaign of 120 runs identified the brightest lead-free composition found in that campaign; the abstract does not report photoluminescence measurements.
- Still unknown: The press release gives no numerical brightness gain; the abstract does not state limitations.
- AI role: Machine learning guided closed-loop experiment selection on a modular microfluidic reactor.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1038/s41467-026-72765-2&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Crystallographic fragment screening of a designed binding protein guided redesign into a fluorophore binder and a Kemp eliminase. LigandMPNN and structure-prediction models supported the design workflow; crystal structures and enzyme assays checked the resulting functions. The paper reports catalytic efficiency of 3.2 million M⁻¹ s⁻¹ for the eliminase. This entry tracks the 4 May 2026 journal version; a preprint preceded it.
Technology: LigandMPNN with Rosetta relaxation for sequence design; RaptorX and later AlphaFold 3 for structure checking
Family AlphaFold 3: profile · models page
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Fragment screening, crystallography and enzyme assays of redesigned proteins; reported Kemp eliminase catalytic efficiency 3,200,000 M⁻¹ s⁻¹.
- Still unknown: The reported kinetics and structural tests are within the study; independent replication was not assessed. The journal version follows an earlier preprint.
- AI role: LigandMPNN with Rosetta relaxation designed enzyme sequences; RaptorX and later AlphaFold 3 predictions checked candidate structures.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13245809/ · https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.1038%2Fs41557-026-02125-6%22&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A K-means framework grouped earthquake families by catalog-derived spatial, temporal, magnitude and seismo-mechanical features. Retrospective analyses found highly localized families with higher interaction and strain release before the 2023 Kahramanmaraş, 2009 L’Aquila and 2014 Iquique earthquakes; similar critical categories were absent before the 2016 Amatrice and 2024 Noto events, where no catalog-visible preparatory phase was detected. The work describes a conditional diagnostic, not a general earthquake prediction system.
Technology: Unsupervised K-means categorization of catalog-derived seismic-family features
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A K-means framework grouped earthquake families by catalog-derived spatial, temporal, magnitude and seismo-mechanical features.
- Still unknown: Peer-reviewed retrospective analysis of five large-earthquake sequences. Preparatory phases in Kahramanmaraş, L’Aquila and Iquique had been documented previously; this paper contributes an unsupervised family-level categorization and comparative feature characterization. The authors state that the approach works only when detectable seismic transients occur, and its use for operational forecasting remains prospective; it is not evidence of reliable prospective prediction.
- AI role: Unsupervised machine learning grouped spatiotemporal event families from multiple seismicity features into categories without predefining the categories; the authors identified a category with highly localized, interacting seismicity and increased strain release in cases with documented preparatory phases.
Evidence sources: https://www.nature.com/articles/s41467-026-72279-x
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Using the open-source OpenEvolve LLM-driven evolutionary framework with a custom reward, the authors find exact values Z(11,21,3,3)=116, Z(11,22,3,3)=121 and Z(12,22,3,3)=132, and lower bounds for 41 further parameter sets, at under $30 per case. Caveats: arXiv preprint not peer reviewed; a separate NeurIPS 2026 listing by overlapping authors reports different counts.
Technology: OpenEvolve (LLM-guided evolutionary search, AlphaEvolve-style)
Family AlphaEvolve: profile · models page
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: The exact values of three Zarankiewicz numbers and lower bounds for 41 further parameter sets were obtained by OpenEvolve search with a tailored reward.
- Still unknown: The page does not describe how the exact values or bounds were verified.
- AI role: OpenEvolve, an open-source LLM-based evolutionary framework, searched for mathematical constructions at under $30 per parameter combination.
Evidence sources: https://arxiv.org/abs/2605.01120
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The T16 pipeline processed 83.7 million TESS Cycle 1 full-frame-image light curves and reported 10,091 new transiting-planet candidates, alongside previously known candidates. Radial-velocity follow-up confirmed TIC 183374187 b as a newly identified hot Jupiter, demonstrating that the semi-automated search yielded at least one bona fide new planet. Most entries in the catalog remain candidates rather than confirmed planets.
Technology: T16 semi-automated transit search with random-forest machine-learning vetting
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The T16 pipeline processed 83.7 million TESS Cycle 1 full-frame-image light curves and reported 10,091 new transiting-planet candidates, alongside previously known candidates.
- Still unknown: The 10,091 entries are planet candidates, most unconfirmed. The cited discovery claim here is limited to TIC 183374187 b, which the authors say was confirmed with radial-velocity follow-up; the pipeline is described as machine-learning-assisted, not fully autonomous.
- AI role: Machine-learning-assisted pipeline classified light curves and prioritized possible transit signals; human vetting and Magellan/PFS radial-velocity observations confirmed the highlighted planet.
Evidence sources: https://arxiv.org/abs/2604.18579
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Researchers used computational protein design and generative AI to recode ribosomal proteins in E. coli so they function without isoleucine, then assembled an engineered strain that remained viable and genomically stable over more than 450 laboratory generations. The result shows a reduced amino-acid alphabet can support core translation machinery in this bacterium; the organism still uses isoleucine elsewhere.
Technology: Protein design models including ESM2, MSATr, ProteinMPNN and AfDesign
Family ESM / ESMFold: profile · models page
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Researchers used computational protein design and generative AI to recode ribosomal proteins in E. coli so they function without isoleucine, then assembled an engineered strain that remained viable and genomically stable over more than 450 laboratory generations.
- Still unknown: This is not a whole 19-amino-acid organism: isoleucine was eliminated from the engineered ribosomal proteins, while the rest of the cell still uses it. Growth was slower than wild type according to the paper.
- AI role: The researchers used AI-guided computational design to propose amino-acid substitutions that allowed selected ribosomal proteins to function after removing isoleucine; the strain was assembled and experimentally tested by the team.
Evidence sources: https://magazine.columbia.edu/article/could-life-be-made-simpler
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: other. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A digital model built from 400,603 cells across 61 single-cell datasets used CellOracle gene-regulatory-network inference and in-silico transcription-factor perturbations to nominate regulators of stem-cell-derived human islet differentiation. CRISPRi experiments supported a previously unreported role for STAT1 in exocrine specification: transient STAT1 knockdown reduced exocrine markers, and JAK–STAT inhibition preserved beta-cell markers and improved glucose-stimulated insulin secretion. CRISPRi also supported a stage-dependent role for ZEB1, whose early knockdown impaired endocrine specification and later knockdown reduced off-target serotonergic cells while increasing beta-cell proportion. The date is the first preprint posting; this entry summarizes the later version reviewed in the research record, which reports the specific STAT1 and ZEB1 experiments.
Technology: CellOracle v0.12.0 within a digital model integrating single-cell transcriptomic and chromatin-accessibility data
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: CRISPRi experiments supported a previously unreported role for STAT1 in exocrine specification: transient STAT1 knockdown reduced exocrine markers, and JAK–STAT inhibition preserved beta-cell markers and improved glucose-stimulated insulin secretion.
- Still unknown: This is a bioRxiv preprint (first posted 30 April 2026; the source lists a later version on 17 September 2026), not peer-reviewed at this source date. It integrates existing datasets and reports laboratory validation in stem-cell-derived islet cultures; it does not establish human therapeutic benefit or mature clinical transplantation outcomes. Regulatory-model predictions remain context- and stage-dependent.
- AI role: The integrated cell-state model inferred regulatory networks and performed in-silico transcription-factor knockout simulations to identify candidate lineage regulators and intervention windows. These model predictions guided subsequent CRISPRi and pharmacologic perturbations in stem-cell-derived islet differentiation cultures.
Evidence sources: https://www.biorxiv.org/content/10.64898/2026.04.27.721124v2.full
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
PEPIC is a sequence-based convolutional model that predicts PET-hydrolyzing potential. The study reports prioritizing W2061_PET9 and experimentally showing that purified protein converts soluble BHET into MHET, confirmed by HPLC and mass spectrometry. The experiment used a soluble PET intermediate; it does not establish degradation of polymer PET. Homology screening and human selection also contributed.
Technology: PEPIC sequence convolutional neural network
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Purified W2061_PET9 converted BHET to MHET in an in-vitro assay confirmed chromatographically and by mass spectrometry.
- Still unknown: Polymer PET degradation, exact sequence-level prior characterization and independent replication are not established.
- AI role: PEPIC scores sequences for predicted PET-hydrolytic activity to prioritize a candidate after genomic/homology selection; direct assay validates soluble BHET turnover.
Evidence sources: https://journals.asm.org/doi/pdf/10.1128/aem.01586-25
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 16 Nov 2026. Review scope: Primary publisher PDF indexed full text confirms title/DOI, publication date, AI role and Fig. 6 experimental validation..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Among nearly 2,000 abdominal CT scans originally considered normal, an AI radiomics model identified 73% of prediagnostic pancreatic cancers, with a median lead time of about 16 months. Validation included images from multiple institutions and protocols. This analysis used scans taken before diagnosis; clinical benefit and false positives in practice still need prospective evaluation.
Technology: REDMOD (Radiomics-based Early Detection Model)
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: REDMOD was trained on 969 patients from a multi-institutional cohort and tested on an independent set of 493 patients (63 pre-diagnostic), with external specificity checks on further datasets and a comparison with radiologists; AUC 0.82 and 73.0% sensitivity, versus 38.9% for radiologists.
- Still unknown: The article states that prospective validation in high-risk cohorts is the necessary next step; no explicit limitations section.
- AI role: A fully automated radiomics framework with AI segmentation and an ensemble classifier on a 40-feature signature from standard-of-care CT scans.
Evidence sources: https://newsnetwork.mayoclinic.org/discussion/mayo-clinic-ai-detects-pancreatic-cancer-up-to-3-years-before-diagnosis-in-landmark-validation-study/ · https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1136/gutjnl-2025-337266&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
An LLM agent (Qiushi Discovery Engine) given only the topic optical computing for AI ran an open-ended study of 145.9 million tokens, 3,242 LLM calls and 1,242 tool calls, and proposed an optical bilinear interaction analogous to the attention operation in Transformers, which the authors say they experimentally validated. The abstract gives no accuracy figure or number of experimental runs. This is an arXiv preprint and the validation is reported by the authors themselves.
Technology: Qiushi Discovery Engine (dual-layer LLM agent with Meta-Trace memory)
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: The abstract says the proposed optical bilinear interaction was experimentally validated on a real optical platform; it gives no measurement details, trial counts or statistics.
- Still unknown: The abstract gives no validation figures and hedges novelty with 'to our knowledge'; hardware use is framed only as a possible route.
- AI role: An LLM-based agentic system (Qiushi Discovery Engine) ran an open-ended study (3,242 LLM calls, 1,242 tool calls, 163 research notes, 44 scripts) and proposed the mechanism.
Evidence sources: https://arxiv.org/abs/2604.27092
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
AggreProt is a deep neural-network predictor of residue-level protein aggregation propensity. Researchers used its predicted aggregation-prone regions to guide mutations in haloalkane dehalogenase LinB and report reduced aggregation, improved soluble yield and up to a 100% increase in yield over wild type under their assay. This is one engineered-enzyme result; model validation on selected peptides/proteins is ancillary, and the result does not establish industrial biodegradation performance.
Technology: AggreProt deep neural-network predictor
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The paper reports the specific in-vitro protein assay summarized here.
- Still unknown: Clinical or diagnostic deployment and independent replication are not established.
- AI role: The model identifies residue-level aggregation regions; predictions inform mutations that are experimentally assessed for aggregation, solubility and yield.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13328301/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Primary paper indexed full text and publication metadata reviewed..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Machine learning models trained on simulated landscape data flagged 9 of 12 known targets (model 1) and all 12 (model 2, with 686 false positives among 709 remaining predictions) in Kisatchie National Forest, Louisiana; a model trained on real South Carolina tar kilns found 11 of 12. Field visits to 11 targets and augering of 2 found none of the expected tar-kiln features, and the authors concluded the structures more likely relate to a WWII military training site. Evidence caveat: only 12 known targets, and field validation is qualitative.
Technology: Machine learning classifiers trained on simulated archaeological sites (with a real tar-kiln-trained comparison model)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Models trained on simulated objects flagged known tar-kiln-like targets, and 11 targets were visited in the field with 2 augered; no expected tar-kiln features were found, and the authors concluded the features are most likely associated with military training exercises.
- Still unknown: Access to all targets was limited by dense understory, and further historical research is needed to confirm the function of the objects.
- AI role: Mask R-CNN models trained on simulated lidar-embedded objects produced predictions, a rule-based script removed likely natural mounds, and human archaeologists reviewed and chose field targets.
Evidence sources: https://www.cambridge.org/core/product/identifier/S2326376825101307/type/journal_article
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Math Inc's autoformalization model Gauss completed the final stages of the Lean verification of Viazovska's proof that the E8 lattice is the optimal sphere packing in dimension 8, working with the human formalization team. The paper describes the human-AI collaboration and the verification in the Lean kernel, with the milestone dated February 2026 in the abstract. Formalization of an existing 2016 result, not a new theorem, and the paper is a progress report.
Technology: Gauss (Math Inc autoformalization model) with Lean
Publication: Preprint · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: prospective or independent check
- Tested: The completed Lean formalization of Viazovska's dimension-8 sphere packing proof was verified in Lean, with the milestone dated February 2026 in the abstract.
- Still unknown: The paper discusses project objectives that remain, which the abstract page does not list; no other limitations are stated.
- AI role: Math Inc.'s autoformalization model Gauss performed the final stages of the Lean verification, working with the human formalization team.
Evidence sources: https://arxiv.org/abs/2604.23468
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 4/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A stacked ensemble trained on 867 compiled degassing records and geospatial raster predictors generated a continuous 0.10° map of model-estimated CO₂ flux within an active-tectonics applicability mask. The authors report spatially coherent high-flux estimates along the Andes, Pacific Ring of Fire and other active orogens. This is a new model-derived spatial estimate for fault-system flux, rather than a direct global survey or measurement.
Technology: Stacked regression ensemble combining support-vector, k-nearest-neighbor, decision-tree, random-forest and gradient-boosting models
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A stacked ensemble trained on 867 compiled degassing records and geospatial raster predictors generated a continuous 0.10° map of model-estimated CO₂ flux within an active-tectonics applicability mask.
- Still unknown: Predictions are model estimates based on 867 compiled records with heterogeneous dates and measurement techniques, not direct measurements at every pixel. The mask is limited to areas within 40 km of mapped active faults or with PGA ≥0.10 g; the 0.10° raster smooths local geology. The abstract reports R²=0.677, while the full text gives stacking test R²=0.693 and validation R²=0.552; fit is moderate. Sampling gaps include central Africa, eastern Asia and North America. High-flux belts overlap known tectonic/geothermal regions, and SHAP associations do not establish causal controls.
- AI role: Five regression learners were tuned and combined in a stacking regressor; that model predicted flux pixel by pixel from global raster inputs and generated the paper’s continuous map of estimated fault-system flux. The mask excluded areas outside the model’s stated tectonically active applicability domain.
Evidence sources: https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2025JH000715
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
OpenAI reports that an internal version of GPT-5.5, with a purpose-built scaffold, found a proof of a known asymptotic result on off-diagonal Ramsey numbers, then verified in Lean. The announcement appeared with the model launch on April 23, 2026; the full proof is in a separate PDF now linked. Page rechecked November 12, 2026: the text matches.
Technology: GPT-5.5 (internal version)
Family GPT-5: profile · models page
Publication: Announcement only · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: prospective or independent check
- Tested: GPT-5.5 found a proof of an asymptotic fact about off-diagonal Ramsey numbers, later verified in Lean.
- Still unknown: Not stated in the reviewed material
- AI role: An internal version of GPT-5.5 with a custom harness found the proof.
Evidence sources: https://openai.com/index/introducing-gpt-5-5/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 4/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A joint molecular model combined bioactivity prediction and molecule reconstruction to estimate how unfamiliar a candidate structure was to its training distribution. In a prospective screen of commercial compounds against PIM1 and CDK1, unfamiliarity-informed selection led to seven compounds with low-micromolar potency and limited similarity to training molecules. The authors also evaluated the metric across 33 experimentally labelled bioactivity datasets.
Technology: Joint molecular model (JMM): semi-supervised molecular property-prediction and reconstruction model using a SMILES encoder; reconstruction loss defines the unfamiliarity score
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The authors also evaluated the metric across 33 experimentally labelled bioactivity datasets. The reported wet-lab candidate assays directly test the molecular-model predictions within the study.
- Still unknown: The kinase finding is a prospective in-vitro hit-identification result, not evidence of therapeutic efficacy. Seven low-micromolar hits were reported across PIM1 and CDK1; CDK1 was described as independent of the study’s earlier dataset analysis. The study screened only a small selected set per target after computational ranking. It does not establish the unfamiliarity measure as universally superior or demonstrate animal or clinical benefit.
- AI role: The model used reconstruction-based unfamiliarity alongside predicted bioactivity and prediction uncertainty to rank molecules for a prospective virtual screen. Researchers tested selected commercial compounds against PIM1 and CDK1; seven compounds showed low-micromolar potency. AI supported prioritization and selection, while wet-lab assays established activity.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13109061/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
In a two-center cohort of 364 participants with type 2 diabetes and controls, targeted serum/urine metabolite measurements found a high-serum/low-urine pattern for indoxyl sulfate and related metabolites in early diabetic kidney disease (DKD), consistent with reduced tubular secretion. Computer vision quantified urine foam stability and chromaticity; Random Forest recursive feature elimination selected these image features together with metabolite ratios and clinical variables for the FluxPro-DKD classifier. It distinguished DKD from diabetes without kidney disease with AUC 0.90 in discovery and 0.85 in a geographically distinct external cohort of 82.
Technology: Computer-vision urine phenotyping and Random Forest recursive feature elimination (RF-RFE) for multimodal feature selection
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: In a two-center cohort of 364 participants with type 2 diabetes and controls, targeted serum/urine metabolite measurements found a high-serum/low-urine pattern for indoxyl sulfate and related metabolites in early diabetic kidney disease (DKD), consistent with reduced tubular secretion.
- Still unknown: The external cohort validation was cross-sectional; the paper says longitudinal follow-up for hard kidney outcomes is ongoing. Its projected three-year event rates are theoretical Monte Carlo simulations based on cross-sectional data and are not observed outcomes. The study did not measure proximal-tubule transporter expression or compare directly against KIM-1 or NGAL, and generalization beyond the two Chinese centers and population requires further validation. Metabolite flux supports, but does not prove, tubular transport dysfunction.
- AI role: Computer vision quantified urine foam half-life and chromaticity from standardized sample images. RF-RFE selected eight predictors, including digital urine features and serum-to-urine metabolite ratios, for the fusion classifier distinguishing early DKD from type 2 diabetes without DKD. The paper reports external cohort evaluation. The biochemical interpretation of flux mismatch as reduced tubular secretory capacity is based on measured metabolite ratios and literature-supported interpretation; transporter expression was not measured.
Evidence sources: https://www.frontiersin.org/journals/endocrinology/articles/10.3389/fendo.2026.1802447/full
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Researchers used an LLM to plan iterative composition measurements and experimentally mapped the Co–Al–Ge phase diagram at 900 °C. The domain-specific aLLoyM model directed early measurements into compositionally complex interior regions, enabling the earliest discovery of all three novel phases reported to form only in the ternary system; the general-purpose LLM strategy found more phases in fewer cycles. This is an arXiv preprint and the finding is bounded to this experimental phase-diagram study.
Technology: aLLoyM domain-specific large language model and general-purpose LLM experimental planner
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: Researchers used an LLM to plan iterative composition measurements and experimentally mapped the Co–Al–Ge phase diagram at 900 °C.
- Still unknown: This source is an arXiv preprint, not a peer-reviewed journal publication as of the checked record. The source abstract supports the reported ternary phase-diagram result and three phases described as novel and ternary-only, but does not provide their formulas or independent replication evidence in the abstract. The discovery is a bounded materials result, not a claim that every phase in the full ternary diagram has been exhaustively mapped.
- AI role: LLMs selected compositions for each cycle in a closed loop with high-throughput synthesis and X-ray diffraction phase identification. The original abstract specifically attributes earliest discovery of all three newly reported ternary-only phases to aLLoyM-guided initial measurements; a general-purpose LLM chose a textbook-like measurement strategy that efficiently found a larger number of phases in fewer cycles.
Evidence sources: https://arxiv.org/abs/2604.20304
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
MetGenX, a structure-informed neural-network workflow for generating structures from tandem mass spectra, was applied to untargeted mouse-liver metabolomics and reported two previously uncharacterized metabolites, adenosyl-cysteine and adenosine 5′-diphosphate 5′-inosine (Ap2I). The paper reports validation against chemical standards using retention time and MS2 spectral matches.
Technology: MetGenX structure-informed encoder-decoder neural network with LightGBM reranking
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: MetGenX, a structure-informed neural-network workflow for generating structures from tandem mass spectra, was applied to untargeted mouse-liver metabolomics and reported two previously uncharacterized metabolites, adenosyl-cysteine and adenosine 5′-diphosphate 5′-inosine (Ap2I).
- Still unknown: The original Nature full-text route was inaccessible in this session; the primary-source PubMed article page provides the paper abstract and figure captions describing standard-based retention-time/MS2 matches for the two structures. The finding is metabolite annotation in mouse tissue, not evidence of human biological function. Independent replication was not checked.
- AI role: The model generated and reranked candidate molecular structures from MS2 spectra; researchers applied a multi-step workflow to mouse-liver samples and compared the two reported metabolite candidates with standards. AI proposed annotations, while standard-based chromatographic and spectral matching supported the assignments.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/42009661/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
An interpretable machine-learning and active-learning workflow screened a virtual library of 18,264 candidate passivation molecules and guided synthesis/testing of selected candidates for perovskite solar cells. The newly identified molecule 5-(aminomethyl)nicotinonitrile hydroiodide (5ANI) yielded 0.05 cm² devices at 27.22% power-conversion efficiency (27.18% certified) and 21.4 cm² mini-modules at 23.49%. Under the reported ISOS-L-1I operating test, 5ANI devices retained 98.7% of initial efficiency after 1,200 hours.
Technology: SISSO interpretable symbolic regression and NGBoost active-learning models, coupled to DFT calculations and a Bayesian-optimization/symbolic-regression automated fabrication platform
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: An interpretable machine-learning and active-learning workflow screened a virtual library of 18,264 candidate passivation molecules and guided synthesis/testing of selected candidates for perovskite solar cells.
- Still unknown: This is a result for a specific perovskite composition, cell architecture and passivation/fabrication protocol. The ML regression was trained on only 20 initial known passivator/device datasets; four selected candidates were experimentally synthesized/validated. The reported champion efficiency is a small-area cell result; the mini-module efficiency is lower (23.49%) and laboratory-scale. The 1,200-hour stability result follows the stated ISOS-L-1I protocol and does not establish commercial lifetime. The paper reports high reproducibility from its automated setup, not factory-scale manufacturing. This is a different study and molecule from the 2024 Bayesian-optimized hole-transport materials record (Science DOI 10.1126/science.ads0901): here AI identifies an interfacial passivator 5ANI.
- AI role: The team used DFT descriptors and SISSO regression trained on measurements from 20 known passivator/device sets to learn structure–performance relationships. NGBoost active learning reduced the computational cost of estimating interfacial properties across an 18,264-molecule virtual library; model-guided screening selected four candidates for synthesis and experimental validation. 5ANI emerged as a passivator for which the automated device-fabrication loop optimized deposition and processing settings. The ML models guided material choice, while the closed-loop platform refined device fabrication and generated measured feedback.
Evidence sources: https://www.nature.com/articles/s41586-026-10482-y
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
An April 2026 arXiv preprint reports 352 lithium-halide spinel samples, covering72% of 171 metal pairs drawn from 19 metals. The proportion meeting conductivity above 0.05 mS/cm and high spinel purity rose from 1.33% in the first 75 agent-proposed samples to 5.33% in the final 75. LLM agents selected experiments; characterization used X-ray diffraction and impedance spectroscopy with manual handling steps. These are within-study laboratory results.
Technology: A-Lab GPSS (LLM agent integrated into a robotic synthesis platform)
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: Robotic synthesis with semi-automated XRD and impedance measurements.
- Still unknown: Not assessed: supplements and scalability outside this campaign. Manual sample handling remains in the characterization workflow.
- AI role: LLM agents used abductive and inductive reasoning for experimental design.
Evidence sources: https://arxiv.org/abs/2604.11957 · https://arxiv.org/html/2604.11957v1
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A knowledge-enhanced AI workflow led to Co22Cr24Fe20Ni25Mo5Mn4, a non-equiatomic high-entropy alloy with experimentally observed single-phase FCC structure and self-limiting oxidation kinetics at 900 °C. After about 50 hours, the study reports a steady-state oxidation-rate constant of 1.41 × 10⁻⁸ mgⁿ cm⁻²ⁿ h⁻¹ (reaction order n = 45.35) and a Cr₂O₃/(Mn,Fe)Cr₂O₄ bilayer oxide scale. The proposed barrier–buffer explanation is the authors’ mechanistic interpretation.
Technology: Physics-guided ML phase classifiers and LLM-assisted alloy-design workflow
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: A knowledge-enhanced AI workflow led to Co22Cr24Fe20Ni25Mo5Mn4, a non-equiatomic high-entropy alloy with experimentally observed single-phase FCC structure and self-limiting oxidation kinetics at 900 °C.
- Still unknown: The specific oxidation result is experimentally measured for one designed alloy under the paper’s 900 °C test conditions and time interval; it is not evidence of service lifetime or performance in a component. The bilayer-oxide barrier–buffer mechanism is the authors’ interpretation from characterization and kinetic behavior. The two additional non-equiatomic FCC/BCC alloys in the study validate phase predictions and remain bundled with the principal material result. The LLM reasoning-agent model is not identified clearly in the paper; DeepSeek-R1-671B is explicitly named for processing-text parsing.
- AI role: The authors used random-forest and support-vector classifiers trained on physics-based descriptors to prioritize single-phase alloy compositions. DeepSeek-R1-671B parsed literature processing descriptions into structured features. Separately, their LLM-agent workflow synthesized oxidation literature and mechanism knowledge to propose composition changes, including increasing Cr and adding about 4 at.% Mn; this guided selection of Co22Cr24Fe20Ni25Mo5Mn4 for synthesis and oxidation experiments. The paper does not clearly identify the model used for the reasoning-agent stage, so that identity is left unspecified.
Evidence sources: https://www.nature.com/articles/s41524-026-02071-4
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Five machine-learning classifiers integrated clinical and serum-proteomic data from carotid atherosclerotic stenosis (60 participants), dyslipidemia (55) and healthy controls (66). A consensus approach identified a six-protein panel (B2M, GPV, MMP9, PLF4, TSP1 and fibrinogen/FB isoforms) that distinguished the atherosclerosis group from the comparison groups, with reported ROC-AUC above 0.9 across the tested classifiers. Protein levels were evaluated by flow cytometry and ELISA in an independent external cohort that included patients with acute atherothrombotic stroke.
Technology: Five machine-learning classifiers: Naive Bayes, linear SVM, random forest, Extra Trees and AdaBoost
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A consensus approach identified a six-protein panel (B2M, GPV, MMP9, PLF4, TSP1 and fibrinogen/FB isoforms) that distinguished the atherosclerosis group from the comparison groups, with reported ROC-AUC above 0.9 across the tested classifiers.
- Still unknown: The discovery cohorts were small (60 atherosclerosis, 55 dyslipidemia and 66 healthy controls); external validation supports the measured panel but does not establish population-wide performance, prospective clinical utility, or improved patient outcomes. The proteins have prior biological associations, so the finding is the specific integrated panel and cohort discrimination rather than discovery of six new proteins or a new mechanism.
- AI role: Five classifiers integrated clinical and serum-proteomic features and selected a consensus six-protein panel for patient-group discrimination. The authors then evaluated the selected proteins in an independent clinical cohort using flow cytometry and ELISA.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/41961364/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A CALPHAD-informed transfer-learning and multi-objective optimization workflow selected two laboratory alloys. The primary abstract reports tensile strengths of 472 ± 7 MPa and 569 ± 12 MPa, with elongations of 23.5 ± 0.5% and 14.9 ± 0.3%. These are measured alloy properties, not industrial qualification.
Technology: CALPHAD-informed transfer learning; ensemble regression; NSGA-II
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The primary source reports the study-specific result summarized here.
- Still unknown: Independent replication and broad deployment are not established by this review.
- AI role: A thermodynamics-informed representation transfers knowledge from established alloy systems; ML predictions and NSGA-II guide experiments, and selected alloys are synthesized and tested.
Evidence sources: https://doi.org/10.1038/s41524-026-02073-2
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Primary-source indexed article or abstract and bibliographic metadata reviewed; complete source scope is given in source_checks..
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A stability-based feature-selection workflow combining LASSO, random forest and SVM recursive feature elimination identified serum IL-8 and CCL3 as a two-protein signature distinguishing Stage I silicosis from controls. The study used a discovery cohort of 57 cases and 57 matched controls, then assessed the signature in an independent, unmatched cohort of 40 cases and 40 controls. The article reports validation AUC 0.973, specificity 95.0% and sensitivity 77.5%; both proteins were lower in silicosis cases.
Technology: Stability-based ensemble feature selection using LASSO, random forest and SVM recursive feature elimination, followed by logistic regression
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: The study used a discovery cohort of 57 cases and 57 matched controls, then assessed the signature in an independent, unmatched cohort of 40 cases and 40 controls.
- Still unknown: The discovery cohort was small (57 cases/57 matched controls); the validation cohort was also small and unmatched (40/40), with 77.5% sensitivity despite high specificity. This is retrospective biomarker validation and does not establish population screening performance, clinical utility, or causal involvement of IL-8 or CCL3 in silicosis.
- AI role: Across 100 iterations, the machine-learning feature-selection framework selected a stable two-protein signature from targeted serum proteomics. A logistic-regression classifier used those selected features and was evaluated by cross-validation and an independent validation cohort. The study reports reduced IL-8 and CCL3 as the signature components.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/41937675/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
An informal reasoning agent (Rethlas) and a formal verification agent (Archon) resolve an open problem in commutative algebra posed by D. D. Anderson in 2014, as a counterexample. The Lean 4 formalization has about 19,450 lines and was produced in roughly 80 agent hours, with the code public on GitHub. Caveat: arXiv preprint; the Lean check covers the formal statement as written.
Technology: Rethlas (informal agent) with Archon (formal Lean agent)
Publication: Preprint · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: prospective or independent check
- Tested: The informal proof was formalized and machine-checked in Lean 4 for the formal statement as written.
- Still unknown: The page describes the case studies as illustrations of capabilities rather than broad evaluations.
- AI role: Rethlas, an informal reasoning agent, built candidate proofs; Archon, a formal verification agent, translated arguments into Lean 4 projects.
Evidence sources: https://arxiv.org/abs/2604.03789
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 4/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A weakly supervised deep-learning analysis combining satellite-derived gravity and bathymetry produced a global seamount inventory and reported systematic morphology differences: broad edifices along fast-spreading ridges and steeper forms near intraplate hotspots. The authors interpret these differences as reflecting geodynamic controls on volcanic construction. Many detections rely on indirect gravity/bathymetry data rather than shipborne mapping, so individual feature identifications and causal interpretations require further validation.
Technology: Weakly supervised deep-learning detector with vision foundation model, iterative pseudo-label refinement, and elliptical Gaussian fitting
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A weakly supervised deep-learning analysis combining satellite-derived gravity and bathymetry produced a global seamount inventory and reported systematic morphology differences: broad edifices along fast-spreading ridges and steeper forms near intraplate hotspots.
- Still unknown: The publisher abstract describes numerous previously undocumented features but does not establish shipborne confirmation for every mapped detection. The reported morphology-setting relationship is an observational pattern and should not be read as proof of a unique causal mechanism.
- AI role: Detected and morphologically characterized seamount features in global gravity and bathymetry data despite sparse labels, enabling the cross-setting comparison that revealed tectonically organized morphology patterns.
Evidence sources: https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2025JH000848
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
DefensePredictor, a machine-learning classifier using protein-language-model embeddings, nominated candidate defense genes across diverse E. coli strains. Of 94 predicted gene clusters selected for testing, experiments validated 42 previously unknown systems; analyses of 1,000 prokaryotic genomes also revealed thousands of uncharacterized protein clusters without homology to known systems.
Technology: DefensePredictor, gradient-boosted classification over protein-language-model embeddings and genomic context
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Of 94 predicted gene clusters selected for testing, experiments validated 42 previously unknown systems; analyses of 1,000 prokaryotic genomes also revealed thousands of uncharacterized protein clusters without homology to known systems.
- Still unknown: The 42 validated systems came from a selected subset of predictions, not all computationally proposed systems; genome-wide counts are predicted protein clusters, not experimentally tested systems. The source reports a 45% validation rate for 94 tested units.
- AI role: The machine-learning model identified likely defense proteins beyond conventional defense-island co-localization methods, guiding experimental cloning and phage-challenge validation. Validation confirmed 42 new systems among 94 tested candidates.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13092281/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Contrastive learning on DESI Legacy Survey DR10 multiband images, followed by SED-fit prioritization, selected sources for spectroscopy; follow-up confirmed 16 quasars at z=5.94–6.45. Several show uncommon narrow Lyα lines, strong Lyα+N V equivalent widths or mildly red observed-frame NIR slopes, and three lie in regions traditional color cuts would miss. The 16-object reported sample includes 15 new literature discoveries and one independently confirmed DESI DR1 source; the 45% follow-up yield includes previously known quasars among 40 observed targets.
Technology: Self-supervised contrastive-learning neural network applied to DESI Legacy Survey DR10 multiband images, with SED-fit prioritization
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Contrastive learning on DESI Legacy Survey DR10 multiband images, followed by SED-fit prioritization, selected sources for spectroscopy; follow-up confirmed 16 quasars at z=5.94–6.45.
- Still unknown: Only 40 candidates were followed up from 1,139 high-priority candidates, so the reported 45% follow-up yield is not completeness or precision over the full sample. That yield counts 18 quasars among 40 targets, including previously known recoveries. The publisher abstract calls all 16 new; the paper body distinguishes 15 new literature discoveries and one additional candidate independently confirmed in DESI DR1 but not reported in literature at the time. Do not overstate population completeness.
- AI role: Learned an image embedding where candidate quasars formed overdensities, allowing atypical sources beyond standard color selection to be prioritized for spectroscopic validation. Spectroscopy confirmed the quasars; AI did not determine their redshifts alone.
Evidence sources: https://www.aanda.org/articles/aa/full_html/2026/04/aa57039-25/aa57039-25.html · https://arxiv.org/html/2603.08830
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The authors prove f(m) = min(m, ceil(2 sqrt(m))) for all m, resolving Erdős Problem #650 on Bloom's site, where f(m) counts disjoint divisibility pairs in an interval for an m-element set. The detailed argument was formally verified in Lean by Harmonic's Aristotle, following a strategy proposed by ChatGPT. Caveat: arXiv preprint; the AI contribution is the strategy and formalization, not an independently checked autonomous solution.
Technology: ChatGPT (strategy) and Aristotle (Lean formalization)
Publication: Preprint · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: prospective or independent check
- Tested: The detailed argument for f(m) = min(m, ceil(2 sqrt(m))) was formally verified in Lean by Harmonic's Aristotle.
- Still unknown: The abstract lists no explicit limitations; the authors state the exposition and final proofs are entirely human-written.
- AI role: ChatGPT proposed the proof strategy and Aristotle produced the Lean formalization.
Evidence sources: https://arxiv.org/abs/2603.28636
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 4/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The study combines trRosetta-based scaffold hallucination and Rosetta design with ProteinMPNN/LigandMPNN sequence design and AlphaFold structural filtering to create NTF2-like binders for six small-molecule targets. An optimized cortisol binder had a measured dissociation constant of 68 nM. A purified-protein, split-luciferase cortisol sensor showed an estimated response EC50 of 72 nM; this is a sensor response measurement, not the binding-affinity value or a clinical detection limit. The sensor also responded to cortisone and 17-alpha-hydroxyprogesterone, so specificity needs improvement. The March 2026 peer-reviewed paper is the source of record.
Technology: trRosetta-based hallucination; Rosetta; ProteinMPNN/LigandMPNN sequence design; AlphaFold structural filtering
Family AlphaFold: profile · models page
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Designed binders for six small-molecule targets were measured for binding affinity (nanomolar to low micromolar), and a cortisol binder was used to build a biosensor.
- Still unknown: The abstract states no specific limitations.
- AI role: Deep learning combined with physics-based methods generated a protein family with varied designable pocket geometries.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1038/s41467-026-70953-8&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A two-stage framework combines deep-kernel active learning with transfer to expanded formulation spaces. The primary article reports electrolytes that increased average lifetime of lithium-metal symmetric cells threefold after three learning iterations and improved formulations in lithium-metal full cells. These are study-specific battery tests, not commercial-pack or independent replication evidence.
Technology: Deep-kernel active learning; target statistic coding; knowledge transfer
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The primary source reports the study-specific result summarized here.
- Still unknown: Independent replication and commercial deployment are not established.
- AI role: Deep active learning selects informative experiments and models formulation performance; statistic coding transfers learned information to expanded composition spaces.
Evidence sources: https://www.nature.com/articles/s41467-026-70973-4
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Primary article/preprint text and metadata as listed in source_checks..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
An LLM-based Catalysis AI Agent identified an energy descriptor for classifying copper-based single-atom alloys and a structural descriptor for predicting activation energy. The team reports experimental testing and that the approach extends to other metal dopants. Evidence caveat: the press release gives no quantitative activity or selectivity values and does not name the experimental group, so validation is limited.
Technology: Catalysis AI Agent (LLM-based agent)
Publication: Announcement only · Test context: Test context unspecified in annotation. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: not reported
- Tested: The press release says the approach was tested experimentally and extends to other metal dopants, but the experiments are not described.
- Still unknown: Not reported; the release gives no quantitative activity or selectivity values and names no experimental group.
- AI role: An LLM-based Catalysis AI Agent trained on the DigCat database mined trends and identified an energy descriptor and a structural descriptor.
Evidence sources: https://www.wpi-aimr.tohoku.ac.jp/en/achievements/press/2026/20260326_002166.html
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 1/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A machine-learning model that predicted antibacterial activity from biosynthetic-gene-cluster sequences prioritized Amycolatopsis azurea DSM 43854. Bioactivity-guided fractionation recovered dipyrimicins A and B from this strain, and genome analysis linked them to a previously uncharacterized dip BGC with a 76% predicted antibacterial probability; broth-microdilution assays confirmed activity against several test strains.
Technology: Machine-learning predictor of antibacterial activity from biosynthetic gene clusters
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Bioactivity-guided fractionation recovered dipyrimicins A and B from this strain, and genome analysis linked them to a previously uncharacterized dip BGC with a 76% predicted antibacterial probability; broth-microdilution assays confirmed activity against several test strains.
- Still unknown: Dipyrimicins themselves had been reported previously from a related Amycolatopsis strain; the paper’s new result is their isolation from A. azurea DSM 43854 and first linkage to its BGC, not discovery of the molecules as novel compounds. Reported antibacterial potency was moderate and tested against a few strains (for dipyrimicin A, MICs included 32 µg/mL for B. spizizenii and 64 µg/mL for S. aureus; >128 µg/mL for P. aeruginosa). Proposed functions of some tailoring genes remain speculative.
- AI role: The model ranked biosynthetic clusters and prioritized A. azurea DSM 43854 for experimental work. Researchers performed bioactivity-guided fractionation, isolated the compounds, linked them to the previously uncharacterized dip BGC, and tested antibacterial activity. The dip BGC was absent from the model’s training set, providing an out-of-training-set test of the prioritization approach.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13122635/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A transformer trained to predict minimum Hamming distance, paired with a genetic algorithm and exact Brouwer–Zimmermann distance checks, found at least six previously unknown champion generalized toric codes over F₈ at dimensions 18, 22 and 38 (length 49). The article reports more than 500 champion codes in its F₈ search and up to a twofold reduction in exact distance evaluations versus random search. The six records at the three newly improved parameter settings are one paper-level coding-theory result.
Technology: Transformer model adapted from GPT-2, coupled with a genetic algorithm
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A transformer trained to predict minimum Hamming distance, paired with a genetic algorithm and exact Brouwer–Zimmermann distance checks, found at least six previously unknown champion generalized toric codes over F₈ at dimensions 18, 22 and 38 (length 49).
- Still unknown: The codes are newly discovered mathematical constructions in the generalized toric family, with minimum distances verified computationally using established exact algorithms and compared to known records. The reported >500 total includes known/previous champion codes, while six at dimensions 18, 22 and 38 are identified as novel; these are not six separate catalog entries. The method searches a restricted family and does not establish optimal codes over all linear codes or quantum codes.
- AI role: The transformer predicts generalized toric codes’ minimum Hamming distances and guides genetic search toward promising generator matrices. Candidate champions are then checked with the Brouwer–Zimmermann algorithm and compared with known bounds; the paper reports six previously unknown champion codes at F₈ dimensions 18, 22 and 38.
Evidence sources: https://www.nature.com/articles/s44387-026-00077-3
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
An ensemble machine-learning framework combined data augmentation, multilevel structural descriptors and physics-informed features to screen more than 10,000 covalent organic framework candidates. The researchers synthesized and tested eight selected COFs. The paper reports COF-343 at 12,978.7 μmol H2O2 h−1 g−1, with additional measured results for COF-2449, COF-2835 and COF-8484. This is one study-level materials result, not four discoveries. Detailed standard-test conditions in the supplement were not reviewed, and the results do not establish industrial performance. A data/code deposit may predate the journal article, but its date and result content were not checked here.
Technology: Deep neural network ensemble; random-forest band-position model
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The indexed primary article reports synthesis and photocatalytic testing of eight selected COFs, including measured H2O2 production rates for COF-343, COF-2449, COF-2835 and COF-8484.
- Still unknown: Detailed standard-test conditions in the supplementary information were not reviewed; the reported laboratory rates do not establish industrial performance. Zenodo deposit chronology/content was not checked.
- AI role: A DNN ensemble predicted photocatalytic yields and a random-forest model estimated band positions to guide COF screening.
Evidence sources: https://www.nature.com/articles/s44160-026-01037-0
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 16 Nov 2026. Review scope: Primary Nature Synthesis article page, abstract, methods/results text and author metadata as indexed by publisher..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Physicist Matthew Schwartz (Harvard) guided Claude Opus 4.5 through a resummation calculation for e+e− collisions, producing a paper with a new factorization theorem and testable predictions against data in about two weeks. Schwartz checked every step; the model claimed checks it had not done and made errors.
Technology: Claude Opus 4.5
Family Claude: profile · models page
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: Results are validated against EVENT2 Monte Carlo simulations, with analytic curves matching the histograms; the paper does not describe a comparison with experimental data, and the author checked each step, with GPT and Gemini cross-checking the written calculations.
- Still unknown: The paper does not describe comparison with experimental data (the author mentions a follow-up project for this); the author states that LLMs cannot yet do original theoretical physics autonomously and that the work needed expert evaluation.
- AI role: Claude Opus 4.5, run through Claude Code, performed the calculations including the factorization derivation, numerical analysis, simulations and manuscript preparation, under physicist supervision.
Evidence sources: https://www.anthropic.com/research/vibe-physics · https://arxiv.org/abs/2601.02484
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A pipeline combining box least squares detection with machine-learning classifiers, trained on realistic simulations, searched 2.2 million TESS main-sequence stars in full-frame images (sectors 1 to 55). It newly validates 118 planets, 31 of them newly detected, and vets over 2,000 candidates, about 1,000 of them new. The machine-learning classifier is a validation tool rather than the detection method, and the paper is a preprint accepted for MNRAS; the peer-reviewed publication date was not confirmed.
Technology: RAVEN: box least squares detection with machine-learning vetting classifiers trained on simulations
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: Candidates from 2.2 million TESS main-sequence stars were classified by machine-learning models trained on realistic simulations and then statistically validated, yielding 118 validated planets.
- Still unknown: Over 2,000 candidates remain unvalidated and may be false positives; the peer-reviewed publication date was not confirmed.
- AI role: Machine-learning classifiers separated planets from false positives; they are a validation tool, not the detection method.
Evidence sources: https://arxiv.org/abs/2603.22597
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A Claude Code agent pipeline analyzed archived 1994 ALEPH hadronic-Z data and reported the first primary Lund jet-plane density measurement in e⁺e⁻ collisions. Its unfolded 57-bin fiducial result shows the expected triangular QCD radiation structure and a shallow low-transverse-momentum turnover; the integrated mean primary-emission count is 4.751 ± 0.224 per hemisphere. This is one new measurement, not a claim of discovering new particles or physics beyond QCD.
Technology: Claude Code agents orchestrated by the Just Furnish Context (JFC) framework
Family Claude: profile · models page
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: A Claude Code agent pipeline analyzed archived 1994 ALEPH hadronic-Z data and reported the first primary Lund jet-plane density measurement in e⁺e⁻ collisions.
- Still unknown: This is an arXiv preprint, and the authors explicitly qualify the novel measurement as subject to independent verification. It uses 1994 ALEPH data and one available full-simulation sample; the corrected density is prior/model dependent, with a reported 6.0% density-weighted total uncertainty and 4.7% uncertainty on the integrated count. The shallow residual shape is within current generator-theory uncertainty, so the paper frames it as a benchmark rather than evidence excluding generators.
- AI role: Agents planned and executed the event selection, jet declustering, unfolding, uncertainty propagation, statistical checks and report drafting using archived ALEPH data and literature retrieval. The paper attributes the new Lund-plane measurement to an autonomous AI-agent analysis pipeline; human authors supplied the broad task and review framework.
Evidence sources: https://arxiv.org/html/2603.20179v3
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Using 138 fb⁻¹ of 13 TeV proton–proton data, CMS reports the first measurement of the W-boson mass in an all-jets final state at a hadron collider: 80.83 ± 0.55 GeV. The result comes from unfolded groomed jet-mass distributions for boosted hadronic W+jets events. The mass-decorrelated ParticleNet graph neural network tagger provides the main result by selecting the W-jet sample and suppressing quark/gluon multijet background; a non-neural N₂ tagger is used as a cross-check.
Technology: ParticleNet mass-decorrelated graph neural network jet tagger
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Using 138 fb⁻¹ of 13 TeV proton–proton data, CMS reports the first measurement of the W-boson mass in an all-jets final state at a hadron collider: 80.83 ± 0.55 GeV.
- Still unknown: This is an all-jets boosted-W extraction with 0.55 GeV total uncertainty, much less precise than the CMS leptonic-channel W-mass result; the paper frames it as a first step toward future all-jets precision measurements and as a test of jet-substructure modeling. The earlier preliminary CMS result of 80.77 ± 0.57 GeV was superseded by this same paper’s published 80.83 ± 0.55 GeV result and is not counted separately. Jet tagging, QCD background estimation, unfolding, hadronization and jet-energy modeling contribute analysis dependence and uncertainties.
- AI role: The analysis uses ParticleNet as its main W-jet tagger. Its graph neural network treats jet constituents as an unordered particle cloud and classifies W→qq′ jets against QCD multijet backgrounds; its mass-decorrelated output defines the signal/control regions used for the primary measurement. The N₂ substructure observable is a separate cross-check.
Evidence sources: https://cms-results.web.cern.ch/cms-results/public-results/publications/SMP-24-012/index.html
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Fritz Haber Institute, with BASF and BasCat, used adaptive experiment planning in a self-driving lab to search over 10^13 promoter combinations with fewer than 50 experiments. The method was validated on propane-to-propylene conversion, and the authors report synergistic promoter effects that explain the catalyst's performance. The claim of outperforming the industry reference is not quantified in the release, so check the paper for numbers.
Technology: Adaptive experiment planning (Bayesian-style, gray-box) in a self-driving lab
Publication: Announcement only · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: direct test
- Tested: The method was validated on propane-to-propylene conversion, and the release says it identified a catalyst superior to the industry reference, without figures.
- Still unknown: No quantitative results are given and the release states no explicit limitations.
- AI role: Adaptive AI planned experiments in a self-driving laboratory, using a gray-box strategy that also explained the effects of individual promoters.
Evidence sources: https://www.fhi.mpg.de/2217409/2026-03-24_ACS_Catal_SDLs
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Tencent researchers used ontology-conditioned generation and experimental feedback to engineer enzymes. In the reported assays, lysozyme TX-RL15 exceeded the earlier TX-L6 design by 66-fold and hen egg-white lysozyme by over two orders of magnitude. Chitinase TX-SC2 retained activity after incubation at 85 °C; selected proteins combined lysozyme and chitinase activity. These are bounded biochemical results. The date represents initial journal publication; the version of record followed on May 7, 2026.
Technology: ORI; protein generation model (PGM); unified sequence model (USM)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Expression, enzyme-activity and heat-treatment assays on selected generated proteins.
- Still unknown: Separate supplements and independent replication not assessed. Assay-specific advantages do not establish industrial or clinical outcomes.
- AI role: Ontology-conditioned generation and reinforcement learning from experimental feedback.
Evidence sources: https://www.nature.com/articles/s41467-026-69855-6?error=cookies_not_supported
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Researchers used the reaction–protein dual-encoder model Horizyn-1 to prioritize enzyme candidates, then experimentally validated activity in orphan-reaction deorphaning, enzyme-promiscuity tests and nonnatural transformations. For an uncharacterized Thermus aquaticus protein, LC–MS confirmed dCMP kinase activity; a selected transaminase produced noncanonical amino acids from lysine-derived donors, with isotope tracing supporting predominant alpha-amine transfer.
Technology: Horizyn-1
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Researchers used the reaction–protein dual-encoder model Horizyn-1 to prioritize enzyme candidates, then experimentally validated activity in orphan-reaction deorphaning, enzyme-promiscuity tests and nonnatural transformations.
- Still unknown: Evidence is from specific in-vitro assays and tested sequences/reactions; model retrieval performance and the tested enzyme activities do not establish broad generalization. The paper reports that the predicted CMP kinase activity for one protein was not confirmed, while dCMP activity was.
- AI role: A contrastive model jointly embedded reaction fingerprints and ProtT5 protein-sequence representations, ranked candidate enzyme–reaction pairs, and informed experimental selection. The authors assayed predictions with biochemical assays, LC–MS and isotope tracing; some fine-tuning results are predictive-performance findings rather than biological discoveries.
Evidence sources: https://doi.org/10.1073/pnas.2520070123
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A multiscale graph embedding workflow prioritized LC-61, which the study experimentally evaluated against L. infantum. The finding is a specific in-vitro compound result; it does not establish clinical efficacy, safety or a validated target.
Technology: Multiscale-aware graph embedding and machine learning
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Primary article/PubMed and institutional manuscript identify LC-61 and its experimental parasite assay.
- Still unknown: Exact prior structures and broader dose/replication details remain open.
- AI role: Graph embedding ranks candidate molecules for assay; human selection and experimental testing follow.
Evidence sources: https://doi.org/10.1021/acs.jcim.5c02947
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: author-institution primary full-text PDF; PubMed primary metadata. Material checked for the specific claims stated in the record..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
ChatHEA, a domain-specific LLM assistant for high-entropy alloy electrocatalysis, guided synthesis and screening of 100 five-element alloy catalysts. The FeCoCuPtIr catalyst reached a peak power density of 0.789 W cm-2 in a fuel cell and outperformed commercial Pt/C in electrochemical tests. These figures are reported by the institutional press release; its numeric activity values and DOE target are not given. The catalog date is the paper's 14 March 2026 publication date, stated in the 8 July press release; the full paper was not reviewed.
Technology: ChatHEA (LLM-based domain assistant)
Publication: Announcement only · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: direct test
- Tested: 100 five-element alloy catalysts were screened by high-throughput experimentation; FeCoCuPtIr outperformed commercial Pt/C in electrochemical tests and reached 0.789 W cm-2 peak power density in a fuel cell.
- Still unknown: Not reported; the page gives no detailed durability data and lists no limitations.
- AI role: ChatHEA, a domain-specific LLM assistant, supported literature review, element design, experiment planning, data processing and activity analysis.
Evidence sources: https://www.wpi-aimr.tohoku.ac.jp/en/achievements/press/2026/20260708_002238.html
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
EquiformerV2-based interatomic predictions screened 3,976 configurations and guided W1–NiFeOOH selection. The journal reports 13.1 A cm−2 at 2.0 V and 80 °C in one anion-exchange-membrane electrolyzer test, and a separate >500-hour test at 1 A cm−2 and 60 °C. Conditions are distinct laboratory tests and are not evidence of commercial scale.
Technology: EquiformerV2 interatomic machine-learning potential
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Publisher primary indexed article reports tested W1–NiFeOOH, high-current-density and separate durability results.
- Still unknown: A dataset deposited 2026-01-10 predates the paper; its preview was inaccessible, so earliest disclosure of the specific result is unresolved. Training overlap and independent replication remain open.
- AI role: Interatomic model predicts stability/adsorption properties to screen catalyst configurations before synthesis and electrolysis tests.
Evidence sources: https://www.nature.com/articles/s41467-026-68735-3
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Primary source checked for identity, publication date, AI workflow and the bounded experimental result stated..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
AlphaEvolve-generated search algorithms improved bounds such as R(3,13) from 60 to 61, R(4,16) from 170 to 174 and R(4,19) from 213 to 219. The paper does not describe independent verification; search code is public. arXiv preprint, latest version read.
Technology: AlphaEvolve
Family AlphaEvolve: profile · models page
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: Search algorithms evolved by AlphaEvolve improved nine Ramsey lower bounds; the authors report recovering bounds for all known exact Ramsey numbers and matching best known bounds in many other cases. No independent verification is described.
- Still unknown: Not reported
- AI role: LLM-based code mutation agent that produced the search algorithms behind all results.
Evidence sources: https://arxiv.org/abs/2603.09172
The legacy label “Not reported” does not distinguish information absent from reviewed material from information not assessed. It is not a scientific finding of absence.
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A collaborative LLM-plus-symbolic-computation study proved that every n×n Latin square with n ≡ 1 (mod 3) has imbalance at least 4n(n−1)/9. It introduced near-perfect permutations that attain the bound and computationally found independently verified matching examples for every such order from 4 through 52. The universal lower-bound theorem was formally checked in Lean 4.
Technology: Claude Opus 4.5 LLM agent with SageMath, Rust exhaustive solver, simulated annealing and Lean 4 verification tools
Family Claude: profile · models page
Publication: Preprint · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: prospective or independent check
- Tested: A collaborative LLM-plus-symbolic-computation study proved that every n×n Latin square with n ≡ 1 (mod 3) has imbalance at least 4n(n−1)/9.
- Still unknown: This is an arXiv preprint (v2 dated 2026-08-13), not yet peer-reviewed in the sources checked. The universal proven theorem is a lower bound for n ≡ 1 mod 3. Matching constructions, and hence demonstrated tightness, are computationally established only for n=4,7,…,52; the paper does not prove that the bound is attained for every larger order. Lean 4 formalizes the lower-bound inequality, not the finite construction search or universal tightness.
- AI role: The agent identified an all-even row-distance parity pattern in numerical search data, formulated and helped prove the lower bound, and used model review to catch proof errors. SageMath and a Rust solver supported symbolic analysis and enumeration; simulated annealing found matching constructions through n=52. Human strategic direction supplied the decisive pivot to the minimum-positive-imbalance question. Lean 4 formally verifies the lower-bound theorem itself.
Evidence sources: https://arxiv.org/html/2603.08322v2
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 4/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
M3-CAD combines sequence, predicted structural and physicochemical features in a multimodal peptide-design pipeline. The journal study reports in-vitro activity for selected QLX-3DV peptides, including untemplated candidates, against tested bacterial strains; selected leads also received additional host-toxicity and mouse wound-infection assays. These are selected preclinical results, not evidence of clinical efficacy or safety. The article cites an earlier M3-CAD/QL-AMP-1 preprint; this entry is restricted to the reported QLX-3DV study result and does not claim the pipeline itself as new.
Technology: M3-CAD multimodal deep-learning pipeline
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Selected designed peptides were synthesized and tested in vitro; selected leads were also examined in mouse wound infection and host-toxicity assays.
- Still unknown: Clinical efficacy and safety, independent replication and full sequence-level comparison with the earlier M3-CAD preprint are not established here.
- AI role: M3-CAD uses multimodal learned sequence and predicted-structure representations with activity-related objectives to generate and rank peptides for experimental testing.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13185819/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 16 Nov 2026. Review scope: Primary PMC full text including publication metadata, methods and experimental results..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Researchers trained ML models on public drug-development data to prioritize biological properties of microbiome metabolites, then tested predictions prospectively. In vitro experiments identified previously unreported IL-8 secretion stimulation by spermine and spermidine, metabolites previously regarded as anti-inflammatory. The paper reports average accuracy of 92.86% in prospective experiments, while cautioning that sample sizes were small.
Technology: Machine-learning property prediction models implemented in KNIME and trained on public drug-development data
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Researchers trained ML models on public drug-development data to prioritize biological properties of microbiome metabolites, then tested predictions prospectively. The selected metabolites were tested in new in-house assays; this is direct experimental testing within the discovery study.
- Still unknown: The unexpected activity is demonstrated in vitro and does not establish in-vivo host effects. The article explicitly cautions that prospective validation sample sizes were small; reported model metrics vary by task.
- AI role: Predicted and triaged microbiome metabolites for target assays, including IL-8 activation; prospective laboratory assays then established the unexpected cytokine-stimulatory activity of spermine and spermidine.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13059116/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Parallelized hybrid-space Bayesian optimization prioritized metal/ligand carriers and conditions for enzyme immobilization. The primary abstract reports activity recovery of 100% for glucose oxidase, 90% for catalase and 79% for Candida antarctica lipase B across the demonstrated workflow. These are enzyme-specific assay outcomes, not discovery of new enzymes or proof of industrial stability.
Technology: Parallelized hybrid-space Bayesian optimization (PHBO)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The primary source reports the study-specific measured result summarized here.
- Still unknown: Independent replication and clinical or industrial efficacy are not established.
- AI role: The optimizer proposes carrier compositions and experimental conditions, then experimental activity feedback informs later proposals.
Evidence sources: https://www.nature.com/articles/s41467-026-70251-3
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Primary source indexed full text or abstract and publication metadata reviewed; scope stated in source_checks..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Semi-supervised image classifiers trained on Pan-STARRS and DESI imagery surfaced five quasar-lens candidates for Keck follow-up. Spectroscopy confirmed GRALJ140833.73+042229.98 as a previously undiscovered gravitationally lensed quasar: a z=2.998 quasar lensed by a z=0.542 early-type galaxy. Three other targets were quasar-plus-star or quasar-plus-galaxy interlopers, and one remained unresolved.
Technology: Semi-supervised image-classifier pipeline using a convolutional autoencoder with a classifier, plus a convolutional network trained with virtual adversarial training
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Semi-supervised image classifiers trained on Pan-STARRS and DESI imagery surfaced five quasar-lens candidates for Keck follow-up.
- Still unknown: Only one of five observed candidates was confirmed; three were interlopers and one could not be resolved. The successful identification came from a previous model iteration during model development, so this is an AI-assisted discovery with limited follow-up yield, not a prospective performance demonstration.
- AI role: Image-only classifiers prioritized candidate lenses; a prior iteration of the autoencoder-classifier surfaced the object, which Keck spectroscopy subsequently confirmed. The paper notes the observed candidates were selected during model development, not from a strictly prospective frozen-model test.
Evidence sources: https://arxiv.org/abs/2504.00054
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A Gemini Deep Think system, guided by tree search and numerical feedback, derived closed-form spectral coefficients for the integral determining gravitational radiation power from cosmic-string loops, resolving the paper’s stated open calculation. The authors compare the expressions against direct numerical integration over ranges of loop harmonic and geometry parameters, finding close agreement. This is a theoretical/computational result, not a new astronomical detection.
Technology: Google Gemini Deep Think integrated with tree search, executable Python evaluation and high-precision numerical feedback
Family Gemini: profile · models page
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: A Gemini Deep Think system, guided by tree search and numerical feedback, derived closed-form spectral coefficients for the integral determining gravitational radiation power from cosmic-string loops, resolving the paper’s stated open calculation.
- Still unknown: The source is an unrefereed arXiv preprint. The paper provides derivations and numerical cross-checks, not proof-assistant verification or independent replication. The first recurrence attempt contained an error that the later Gemini-assisted review corrected. The claim is limited to the stated integral for the cosmic-string radiation model and its exact analytic evaluation.
- AI role: Gemini generated and refined analytical approaches inside a tree-search workflow whose Python harness compared candidate expressions with high-precision numerical integration. A later human-initiated session with a more advanced Gemini model identified a recurrence error and telescoped the series into a finite analytic expression; this was a human–AI handoff, not a fully autonomous end-to-end result.
Evidence sources: https://arxiv.org/abs/2603.04735
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A time-delay neural network trained on multi-parameter acoustic-emission records from Alzo-granite triaxial tests at 5–40 MPa resolved three recurring phases: microcrack nucleation marked by waveform-scattering changes, fault growth with evolving spatial and magnitude distributions, and coalescence marked by peak-delay increases and b-value changes. The model also predicted stress-drop timing and size across axial-splitting and shear-localization failures. This is a laboratory result, not a natural-earthquake precursor claim.
Technology: Genetic-algorithm-optimized time-delay neural network (TDNN)
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A time-delay neural network trained on multi-parameter acoustic-emission records from Alzo-granite triaxial tests at 5–40 MPa resolved three recurring phases: microcrack nucleation marked by waveform-scattering changes, fault growth with evolving spatial and magnitude distributions, and coalescence marked by peak-delay increases and b-value changes.
- Still unknown: The analysis uses four laboratory experiments in Alzo granite, so results are sample- and apparatus-limited and do not establish natural-earthquake predictability. Training-data smoothing and the small number of experiments constrain generalization; stress estimates show more variability and modelled subtle phase transitions are less accurate. The observed phase interpretation is based on model-assisted analysis of acoustic emissions and measured mechanical data.
- AI role: The TDNN integrated time-dependent acoustic-emission waveform features (peak delay, coda Q/scattering) and occurrence measures (event rate, Gutenberg–Richter b-value, spatial fractal dimension), dynamically weighting them to model stress/strain evolution and identify repeatable phases across pressure conditions. Leave-one-experiment-out validation tested cross-condition generalization.
Evidence sources: https://www.nature.com/articles/s43247-025-03003-8
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A topology-aware generative workflow and activity predictor prioritized quaternary ammonium compounds (QACs) for synthesis and testing. Across two workflows the authors report 29 selected compounds characterized and 11 QACs designated as novel, with measured minimum inhibitory concentrations of 1–32 μM against bacterial pathogens. Earlier same-team conference disclosures had already described AI-generated QAC design and synthesis; this record covers the later paper’s bounded tested results and does not claim the first AI-generated QACs or universal chemical novelty. Activity and cytotoxicity remain assay-specific and do not establish clinical use.
Technology: Topology-aware QAC generative model with structure/activity prediction and filtering (Q-HERO)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The article reports synthesis and antimicrobial testing across two workflows, with 29 selected compounds characterized and 11 reported novel QACs with MIC values of 1–32 μM.
- Still unknown: Prior same-team AI-QAC design and synthesis were described in 2022 and 2024 institutional conference abstracts. The measured activity is limited to the reported laboratory assays and does not establish clinical efficacy.
- AI role: A topology-aware generative model proposed QAC structures, and learned activity filtering helped prioritize candidates for synthesis and antimicrobial assays.
Evidence sources: https://pubs.acs.org/doi/10.1021/acs.jcim.6c00390 · https://pubs.acs.org/jcisd8/article/66/6/3117/5080435/Topology-Aware-Generation-and-Activity-Based
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary ACS indexed full article, publication history, methods/results, named affiliations and indexed supporting-information description; earlier team conference disclosures as identified in the frozen source review..
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Building on work on gluons, GPT-5.2 Pro helped derive formulas for nonzero single-minus gravitational amplitudes in a special semi-collinear regime. The authors report analytical verification and checks against known physical limits; this is a theoretical preprint without experimental confirmation.
Technology: GPT-5.2 Pro
Family GPT-5: profile · models page
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: The preprint derives explicit formulas for single-minus graviton tree amplitudes in the half-collinear regime and shows they follow from symmetry principles and recursion relations; no experimental test is described.
- Still unknown: The page calls it a small step towards reconciling quantum mechanics with general relativity.
- AI role: Researchers used GPT-5.2 Pro to help find the result; the paper is authored by named physicists on behalf of OpenAI.
Evidence sources: https://openai.com/index/extending-single-minus-amplitudes-to-gravitons/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The BarFlare assay and an interpretable multi-criteria decision-making classification-fusion machine-learning framework combined extracellular-vesicle SCC and MMP13 biomarkers with clinical factors. The model was evaluated in prospective multicentre diagnostic cohorts and two population cohorts; the longitudinal cohort contained baseline samples from 30 people who later developed oesophageal squamous cell carcinoma, with a reported median lead time of 34.9 months and AUC 0.864. The paper also reports external diagnostic validation in two cohorts.
Technology: Multi-criteria decision-making Classification Fusion (MCF) machine-learning framework; 289 candidate feature-selector/classifier pairs were compared and three base models fused
Publication: Peer-reviewed paper · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: prospective or independent check
- Tested: The BarFlare assay and an interpretable multi-criteria decision-making classification-fusion machine-learning framework combined extracellular-vesicle SCC and MMP13 biomarkers with clinical factors.
- Still unknown: This is a prospective observational diagnostic study, not an intervention trial or proof that screening improves outcomes. The longitudinal preclinical cohort had 30 incident cases and matched controls; performance estimates require external replication in larger populations. The source reports an association and lead time, not guaranteed individual prediction. Primary publisher full text and abstract checked; no independent replication assessed.
- AI role: The MCF framework integrated newly identified EV biomarkers with clinical variables, selected and fused machine-learning classifiers, and produced the risk score evaluated in diagnostic and prospective population cohorts. BarFlare is the high-sensitivity EV assay platform, not the AI model.
Evidence sources: https://isevjournals.onlinelibrary.wiley.com/doi/10.1002/jev2.70246
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 4/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A protein-language-model pipeline, HMD-AMP, screened host and gut-microorganism genomes from nine mammalian species and nominated antimicrobial peptides, including candidates with little similarity to known AMPs. Of 91 high-confidence peptides selected for experiments, 74 showed strong antibacterial activity and 48 were evolutionarily remote from known AMPs. Four tested peptides had broad-spectrum activity at low effective concentrations and low toxicity; the most potent reported candidate showed efficacy in a mouse peritoneal E. coli infection model.
Technology: HMD-AMP protein-language-model pipeline
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Of 91 high-confidence peptides selected for experiments, 74 showed strong antibacterial activity and 48 were evolutionarily remote from known AMPs.
- Still unknown: These are in-vitro peptide activity results plus a single reported mouse peritoneal E. coli infection experiment; they do not establish human safety or efficacy, clinical antibiotic utility, or activity against every pathogen. Only 91 of the many predicted sequences were tested, and 74 showed strong activity under the authors’ assays. The entry is one paper-level finding; other AMP discovery records in this ledger use different candidate sequences, sources or design workflows. The model acronym overlaps an earlier 2021 HMD-AMP classifier preprint, but this 2026 paper is a separate study with mammalian metagenomic mining and experimental outcomes; they are not conflated.
- AI role: HMD-AMP used protein language-model representations and a hierarchical classifier to identify and rank candidate antimicrobial peptides from host and gut-microorganism metagenomic/genomic sequences. The researchers experimentally tested 91 high-confidence candidates; 74 showed strong antibacterial activity, including 48 evolutionarily remote sequences, and four had the reported broad-spectrum/low-toxicity profile. The strongest candidate was then evaluated in a mouse infection model.
Evidence sources: https://www.nature.com/articles/s41551-026-01630-w
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A CatBoost model trained on CMIP6 simulations used the observed 1971–2020 grid-cell warming pattern to identify tropical and polar regions that best constrain later global warming. Applying these spatial constraints reduced projection error variance by about 70%, compared with about 48% using global-mean warming alone; the results depend on model-based emergent relationships and are projections rather than direct observations of future climate.
Technology: CatBoost machine-learning emergent-constraint model
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A CatBoost model trained on CMIP6 simulations used the observed 1971–2020 grid-cell warming pattern to identify tropical and polar regions that best constrain later global warming.
- Still unknown: The study is a climate-model emergent-constraint analysis; uncertainty reduction was assessed with leave-one-model-out cross-validation, not by validating future outcomes. The reported probabilities are scenario-conditional projections.
- AI role: Learned grid-cell-scale relationships between historical warming fields and future global mean warming across CMIP6 simulations, then ranked grid cells by their influence on constrained projections.
Evidence sources: https://www.nature.com/articles/s41467-026-70205-9
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A University of Hong Kong-led team used a conditional variational autoencoder constrained by biochemical properties and a random-forest activity classifier to generate and prioritize antimicrobial peptides. In vitro testing of 54 generated peptides found 18.5% with MIC ≤10 μg/mL and 38.9% with MIC ≤50 μg/mL against the study’s E. coli model; nine peptides were classified as both active and non-toxic in the reported assays.
Technology: CVAE-BIO: conditional variational autoencoder with biochemical constraints and a random-forest antimicrobial-activity classifier
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: In vitro testing of 54 generated peptides found 18.5% with MIC ≤10 μg/mL and 38.9% with MIC ≤50 μg/mL against the study’s E. coli model; nine peptides were classified as both active and non-toxic in the reported assays.
- Still unknown: Validation is in vitro, primarily against E. coli ATCC 25922; it is not evidence of treatment efficacy in animals or humans. The reported nine active/non-toxic peptides are assay-defined candidate hits, not validated drugs. This paper generated a different peptide set and used different training/design and testing workflows from the earlier C. acnes deep-learning, bullfrog-genome AMPlify, and LSTM Klebsiella sepsis studies in this ledger; do not treat the general AMP topic as a duplicate.
- AI role: The CVAE generated peptide sequences subject to predefined biochemical constraints; the random forest ranked activity using 30 biochemical descriptors. Researchers synthesized/tested 54 generated peptides in vitro, measured MIC and hemolysis, and reported the active/non-toxic subset.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/41849223/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Knuth's note reports that Claude Opus 4.6 found an explicit decomposition of the arcs of the m^3-vertex digraph into three directed Hamiltonian cycles for odd m, an open problem Knuth posed. Knuth's proof of the construction was reportedly formally verified in Lean by Kim Morrison. Caveat: sources disagree on the even case, which an independent arXiv paper (2603.24708) and a GPT-5.4 Pro proof treat differently.
Technology: Claude Opus 4.6
Family Claude: profile · models page
Publication: Not annotated in this catalog · Test context: Not annotated in this catalog. Catalog categories, not an evidence grade.
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: other. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 4/5 · Novelty 5/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A random-forest classifier ranked unidentified Fermi-LAT sources, from which the team selected 79 targets for MeerKAT UHF observations. The survey reported nine new millisecond pulsars, all detected in coherent tied-array beams; seven were also confirmed as Fermi-LAT counterparts through gamma-ray pulsations. This entry covers only the nine MSP discoveries. Six slower pulsars were also reported, but five were found in a wider incoherent beam and are not included in the AI-linked result. The authors note earlier optical redback-candidate identification for PSR J1910−5320 and single-pulse reports in the J0657−4657 beam; no claim is made that these were the first observations at all wavelengths.
Technology: Random Forest classifier used to rank unidentified Fermi-LAT sources as pulsar-like candidates
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The survey observed 79 random-forest-selected Fermi-LAT sources and reports nine new MSPs, all detected in coherent tied-array beams; seven were confirmed as Fermi-LAT counterparts through gamma-ray pulsations.
- Still unknown: The nine-MSP record is limited to the coherent-beam result. The paper also reports six slow pulsars, but five were detected in the wider incoherent beam and are not included in this AI-linked claim. The authors note an earlier optical redback-candidate identification for J1910−5320 and single-pulse reports from the J0657−4657 beam; the entry does not claim priority across wavelengths or treat those reports as prior radio-pulsar discoveries.
- AI role: A random-forest model ranked unassociated Fermi-LAT sources for target selection; radio searches of 79 selected targets yielded the nine coherent-beam MSP discoveries.
Evidence sources: https://arxiv.org/html/2602.20811v1 · https://academic.oup.com/mnras/article/547/4/stag406/8501212 · https://pure.mpg.de/rest/items/item_3695468_3/component/file_3702533/content
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary arXiv v1 full text for model-based target selection, coherent-beam results, affiliations and prior-source caveats; primary OUP indexed journal record and author-deposited peer-reviewed version for publication date and claim confirmation..
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
CMS reports the first observation of single-top-quark production in association with a W and Z boson in proton-proton collisions. A Particle Transformer classifier and improved reconstruction separated the rare tWZ signal from major backgrounds in three- and four-lepton events. Combining 13 and 13.6 TeV data (200 fb⁻¹), the measured signal has 5.8σ observed significance (3.5σ expected) and cross sections of 248 ± 52 fb at 13 TeV and 242 ± 77 fb at 13.6 TeV. The result agrees with the Standard Model within 2.3σ.
Technology: Particle Transformer (ParT) multiclass classifier
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: CMS reports the first observation of single-top-quark production in association with a W and Z boson in proton-proton collisions.
- Still unknown: This is the first observation of the Standard-Model tWZ production process, not evidence for physics beyond the Standard Model. A 2024 CMS analysis had reported earlier evidence at 3.4σ; the 2026 paper adds 13.6 TeV Run 3 data and improved reconstruction/ML analysis to reach observation. The 5.8σ result combines 13 and 13.6 TeV datasets. Cross sections and signal strength are consistent with the Standard Model within the reported uncertainties.
- AI role: CMS trained four ParT models, split by three- versus four-lepton signal region and 13 versus 13.6 TeV samples. In the three-lepton region, a four-node classifier distinguished tWZ, ttZ and other backgrounds; in the four-lepton region, a binary classifier separated tWZ from backgrounds. Its classifier-score categories entered the statistical signal extraction. The paper reports a substantial sensitivity gain versus the previous analysis: expected significance on the reused 13 TeV data rose from 1.4σ to 3.0σ, contributing to the first observation when combined with 13.6 TeV data.
Evidence sources: https://cms-results.web.cern.ch/cms-results/public-results/publications/TOP-24-009/ · https://arxiv.org/abs/2510.19080
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
An iterative machine-learning-guided campaign optimized precious-metal and additive compositions and support over 35 prediction–experiment cycles. The authors report a low-loading Pt–Ru–Pd catalyst that outperformed a commercial benchmark and their earlier high-PGM design in the reported reverse water–gas shift assay. This is a laboratory catalyst result, not demonstrated industrial performance.
Technology: Machine-learning-guided iterative catalyst composition and support optimization
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The publisher article reports the 35-cycle campaign and its best low-PGM catalyst outperforming the stated commercial and earlier high-PGM comparators.
- Still unknown: The result is specific to the reported assay; scale-up and industrial operation are not established.
- AI role: Machine-learning-guided predictions prioritize PGM/additive compositions and support for iterative experimental testing.
Evidence sources: https://pubs.acs.org/doi/10.1021/acscatal.5c09278
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Publisher indexed primary article text including author affiliations, article history and abstract/result; full-text access remains publisher-limited..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
In 406 healthy GEOCODE cohort participants, whole-blood cytokine responses to Toll-like-receptor (TLR) and T-cell-receptor (TCR) stimulation varied across people, including high- and low-IL-6-response profiles associated with monocyte counts, sex and sampling season. Feature-importance analysis of machine-learning models found TCR-induced cytokine predictions were mainly driven by genetic features, whereas TLR-induced predictions were driven more by environmental and biological features. Combining genetic with biological and environmental data improved prediction by about 0.2 on average, and the reported findings were evaluated in the independent Milieu Intérieur cohort.
Technology: Machine-learning models including random forest, gradient-boosted trees, artificial neural networks and elastic net; random forest feature-importance analysis
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: In 406 healthy GEOCODE cohort participants, whole-blood cytokine responses to Toll-like-receptor (TLR) and T-cell-receptor (TCR) stimulation varied across people, including high- and low-IL-6-response profiles associated with monocyte counts, sex and sampling season.
- Still unknown: The discovery cohort was a selected healthy-adult sample (406 participants), and the correlations for TLR-induced response prediction were modest (average 0.27; TCR-induced 0.53). Genetic-only models may overestimate predictive accuracy, as the authors report. Feature importance identifies predictive contribution, not causal biological effects; results do not establish disease-specific immune mechanisms or treatment response.
- AI role: Models predicted measured cytokine levels following TLR or TCR stimulation from genetic, biological and environmental participant features. Random-forest permutation importance identified which input categories contributed to predictions; the study also compared predictions across cross-validation and a separate Milieu Intérieur cohort. Feature importance is associational and does not establish that a feature causally regulates cytokine production.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/41735307/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A fully convolutional network segmented subduction zones in two-dimensional mantle-convection simulations using temperature, vertical velocity and fineness fields. Across the modeled runs, zones forming near continental margins were generally longer-lived and less sensitive to internal heating than oceanic zones, supporting the hypothesis that continental lithosphere can stabilize subduction during early-Earth-like conditions. This is a computational-model result rather than direct observation of ancient Earth.
Technology: Attention-based fully convolutional network with custom segmentation loss
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A fully convolutional network segmented subduction zones in two-dimensional mantle-convection simulations using temperature, vertical velocity and fineness fields.
- Still unknown: The physical inference comes from two-dimensional numerical mantle-convection models and their selected parameter space, not empirical reconstruction of Earth's early subduction history. The AI performs segmentation/tracking; the tectonic contrast is a statistical result from model outputs.
- AI role: Provided threshold-free, spatially contextual tracking of evolving subduction zones in simulation outputs; the resulting tracked histories enabled statistical comparison of lifetimes and sensitivity to mantle convective vigor by tectonic setting.
Evidence sources: https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/2025JB032355
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
ETopt combines sequence, structural and protein-language-model features in an ensemble XGBoost model to predict enzyme optimum temperatures. The authors report wet-lab screening that identified two novel thermophilic beta-agarases. The draft is limited to those source-identified enzyme findings; model benchmarks and other screened enzyme classes are not counted as separate discoveries. Industrial process performance and independent replication are not established.
Technology: ETopt multimodal model; ensemble XGBoost; ProtBERT and Prime embeddings
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The source highlights experimental identification of two novel thermophilic beta-agarases; detailed enzyme activity and comparator data were not available in the reviewed abstract.
- Still unknown: Full article, exact assay results, sequences, prior enzyme overlap and independent replication were not reviewed.
- AI role: ETopt integrates learned sequence and structural representations with selected features to predict optimum temperature and prioritize candidate enzymes for wet-lab validation.
Evidence sources: https://www.sciencedirect.com/science/article/pii/S0141813026009943 · https://pubmed.ncbi.nlm.nih.gov/41747980/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 16 Nov 2026. Review scope: Publisher highlights and abstract; PubMed primary metadata and abstract..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A GPT-4o-based agent (LFAST) screened reaction conditions over a 96-condition matrix and refined them after each round, raising the crystallinity index of a reference covalent organic framework by 350% and yielding COF-2000, a previously unreported material with higher structural order. The reported validation is crystallinity measurement of the synthesized products; the authors do not pursue applications. Publication date is taken from secondary reports of the JACS paper and should be checked against the journal page.
Technology: LFAST (LLM For Accelerated Synthesis Technique), GPT-4o agent with robotic experimentation
Family GPT-4: profile · models page
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Products of LFAST-guided conditions characterized by powder X-ray diffraction: crystallinity index raised by 350% on the benchmark COF TpPa-SO3H; COF-2000 characterized by PXRD and crystallinity index only.
- Still unknown: Evidence rests on one benchmark system and one new COF; the full text was not read, so limitations discussed there are unknown.
- AI role: LLM deep research agent mined the literature for synthesis conditions; an automated platform ran them and PXRD analyzed the products.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1021/jacs.5c23233&resultType=core&format=json · https://bidmap.berkeley.edu/news/new-ai-agent-paradigm-shift-cof-synthesis
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A sequence-to-sequence transformer generated readable Python meta-programs that construct photonic experimental blueprints for entire families of quantum states, rather than only individual target states. In the final paper, two successful generalizations were previously unknown: a photonic construction for general spin-1/2 states with no adjacent spin-ups, and one for states of the Majumdar–Ghosh model. The generated programs matched states beyond the three examples used as input, providing computational evidence of the inferred rules.
Technology: Transformer-based sequence-to-sequence language model trained on synthetic quantum-state and experimental-blueprint pairs
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A sequence-to-sequence transformer generated readable Python meta-programs that construct photonic experimental blueprints for entire families of quantum states, rather than only individual target states.
- Still unknown: This is a computational design result, not a laboratory realization of the proposed photonic experiments. Generalization was tested by checking generated states at finite larger sizes; the authors state that arbitrary code equivalence for all system sizes is not straightforward to prove. The two constructions are counted as one paper-level discovery.
- AI role: Generated human-readable Python meta-programs from short sequences of target quantum states; the model’s programs proposed reusable experimental construction rules for whole state families, which the authors checked by simulating output states at larger system sizes.
Evidence sources: https://www.nature.com/articles/s42256-025-01153-0
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A neural network trained on measured growth responses from 2,560 compounds screened 6,111 FDA-evaluated compounds for activity against multidrug-resistant emm92 group A Streptococcus. Of nine selected compounds tested, GW3965 reduced bacterial survival by 99% at an MIC of 6.25 μM. The authors also report improved wound closure in an infected human skin-equivalent model and reduced lesion area and bacterial burden in mice. This is preclinical evidence and does not establish efficacy or safety in people. The peer-reviewed journal article followed on 2026-05-30 and reports the same core GW3965 result.
Technology: Deep neural network trained on compound SMILES and measured bacterial growth responses; model screened an FDA-evaluated compound library
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: The model was trained on 2,560 compounds and screened 6,111. Nine selected compounds were validated; GW3965 had MIC 6.25 μM and reduced viable emm92 iGAS by 99%. In a human skin-equivalent model infected tissues treated with GW3965 showed wound closure; in a mouse SSTI model 30 mg/kg reduced lesion size and day-6 wound bacterial burden by about 2 logs on average.
- Still unknown: Results come from in-vitro assays and animal/tissue models; the preprint reports no synergy with ampicillin. Human clinical efficacy, dosing, safety and mechanism against GAS remain unestablished.
- AI role: A neural network learned structure–growth-inhibition patterns from measured compound screening data, prioritized GW3965 for follow-up, and the selected candidate was experimentally tested in bacterial, human skin-equivalent, and mouse infection models.
Evidence sources: https://assets-eu.researchsquare.com/files/rs-8779898/v1/a183f808-a19a-41be-8b56-ad4009293888.pdf
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary Research Square v1 full-text PDF, including posted date, affiliations, Methods and experimental outcomes; primary Nature publisher indexed article text and journal-version abstract/results..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A GRU encoder-decoder language model learns host-specific amino-acid-to-codon patterns and designs synonymous sequences. Across six proteins, the study reports the best titer for five against four commercial optimization tools; human serum albumin reached about threefold the native-sequence titer. Results are plate-scale expression measurements, not manufacturing economics or clinical product outcomes.
Technology: Pichia-CLM (GRU encoder-decoder)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The source reports the study-specific result summarized here.
- Still unknown: Independent replication and broader deployment or clinical benefit are not established by this source review.
- AI role: The learned host codon language model generates optimized coding sequences; researchers experimentally compare protein titers with native sequences and four commercial tools.
Evidence sources: https://www.pnas.org/doi/10.1073/pnas.2522052123
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 16 Nov 2026. Review scope: Primary-source abstract and publication metadata reviewed; full text scope is specified in source_checks..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Researchers designed four protein exoframe modulators targeting the transmembrane region of the dopamine D1 receptor. Structural studies and functional assays showed agonist-positive, negative and biased allosteric modulation; an agonist-positive modulator restored activity in loss-of-function receptor mutants. The journal abstract supports these receptor-level experiments, rather than a clinical treatment claim.
Technology: Computational de novo protein design (AI-based protein design pipeline)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Structural and functional investigation of four designed modulators of the D1 receptor, including rescue of loss-of-function mutants.
- Still unknown: Review limited to abstract, online date and affiliations. Therapeutic efficacy in patients is not established by these receptor experiments.
- AI role: A deep-learning, hallucination-like protein-design workflow used structural prompts to generate transmembrane binding modes.
Evidence sources: https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI%3A%2210.1038%2Fs41586-025-09957-1%22&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
GPT-5.2 conjectured a formula for gluon scattering amplitudes with a single negative helicity in a special semi-collinear regime. An internal model produced a proof that the authors analytically verified; the announcement describes a preprint, not experimental confirmation.
Technology: GPT-5.2 Pro
Family GPT-5: profile · models page
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: The arXiv abstract reports that the gluon formula satisfies several consistency conditions, including Weinberg's soft theorem, and that the formula is restricted to half-collinear configurations; no experimental test is described.
- Still unknown: Not reported (the abstract states no limitations beyond the restricted half-collinear and complexified-momentum setting).
- AI role: GPT-5.2 conjectured the formula; an internal model produced a proof that the authors verified analytically.
Evidence sources: https://arxiv.org/abs/2602.12176
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 2/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A MobileMamba-ECA image classifier searched DESI Legacy Imaging Surveys data and surfaced 775 Green Pea galaxy candidates. Cross-matching with DESI DR1 spectroscopy established 216 as previously unreported Green Pea galaxies: although spectra existed, the standard DESI pipeline had classified them only as generic galaxies. The study’s remaining 511 photometric-only objects are follow-up candidates and are not included in the confirmed count.
Technology: MobileMamba-ECA deep-learning classifier (Mamba state-space backbone with efficient channel attention)
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A MobileMamba-ECA image classifier searched DESI Legacy Imaging Surveys data and surfaced 775 Green Pea galaxy candidates.
- Still unknown: The 216 newly identified galaxies have archival spectra and were classified with emission-line and BPT diagnostics; they were not all newly observed. The 511 sources without archival spectral coverage remain candidates and are excluded from this finding. The result is a newly recognized population/sample, not a claim that the galaxies themselves formed recently.
- AI role: Classified multiband DESI Legacy Survey images to prioritize compact Green Pea galaxy sources from an 843,592-object parent sample; this enabled spectral cross-identification and recognition of 216 previously unreported examples.
Evidence sources: https://academic.oup.com/mnras/article/547/1/stag278/8475333
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The Aletheia mathematical agent scored up to 90% on IMO-ProofBench Advanced and independently solved four open Erdős questions; work with experts covered 18 research problems, including a disproof of a submodular optimization conjecture. Its score on doctoral-level problems falls to about 38–46%; no result is claimed as a breakthrough, and the papers were submitted to journals.
Technology: Gemini Deep Think, Aletheia agent
Family Gemini: profile · models page
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: Aletheia was evaluated on IMO-ProofBench Advanced (about 90%, human experts graded), an internal PhD-level benchmark (about 46%), a semi-autonomous run over 700 open Erdős problems (four autonomous solutions), and expert collaborations; Level 2 papers were submitted to journals.
- Still unknown: The article states that no Level 3 or Level 4 results are claimed; the abstract states no limitations, and the PhD-level score falls to about 38-46%.
- AI role: Aletheia, a math research agent built on Gemini Deep Think, generated, verified and revised proofs and used search, with varying autonomy across projects.
Evidence sources: https://deepmind.google/blog/accelerating-mathematical-and-scientific-discovery-with-gemini-deep-think/ · https://arxiv.org/abs/2602.03837
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The CHAI computational-histology platform generated a locked H&E-based biomarker intended to distinguish advanced pancreatic ductal adenocarcinoma patients with differential outcomes on fluoropyrimidine-based versus gemcitabine-based chemotherapy. In an independent validation cohort of 299 patients from the prospectively collected COMPASS and Know Your Tumor cohorts, the 173 patients classified F-pref had longer time to next treatment or death (8.6 vs 7.5 months, P=.035) and overall survival (14.4 vs 11.7 months, P=.003) on F- versus G-based therapy. Among 126 G-pref patients, G-based therapy was associated with longer time to next treatment or death (9.6 vs 7.2 months, P=.038), but not significantly different overall survival.
Technology: Computational Histology Artificial Intelligence (CHAI) platform
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: In an independent validation cohort of 299 patients from the prospectively collected COMPASS and Know Your Tumor cohorts, the 173 patients classified F-pref had longer time to next treatment or death (8.6 vs 7.5 months, P=.035) and overall survival (14.4 vs 11.7 months, P=.003) on F- versus G-based therapy. The locked biomarker was applied to assembled data from prospectively collected cohorts; prospective validation of treatment selection is still needed.
- Still unknown: Validation used 299 patients in two prospectively collected but nonrandomized cohorts; treatment assignment was not randomized, and the authors also used propensity weighting and target-trial emulation. The primary endpoint was time to next treatment or death, not progression-free survival. In the biomarker’s G-pref subgroup, the OS comparison was not significant. The authors call for further prospective validation before clinical adoption. An earlier CHAI study concerned adjuvant gemcitabine outcomes in resected disease; this entry is limited to the distinct locked first-line fluoropyrimidine-versus-gemcitabine biomarker and external-cohort results in advanced disease.
- AI role: Deep-learning models segmented H&E biopsy slides and quantified more than 30,000 histomorphologic features. Using the development cohort, the investigators statistically selected and weighted features into a treatment-preference biomarker and locked its threshold before applying it to an independent validation cohort. The platform supplied the image-derived feature measurements used in the biomarker; this is a biomarker-development and validation result, not autonomous treatment assignment.
Evidence sources: https://ascopubs.org/doi/pdfdirect/10.1200/JCO-25-02199
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A domain-tuned QwQ-32B model generated precursor-additive suggestions. Researchers tested selected recommendations in perovskite solar-cell devices against controls and manually selected additives and report improved device performance and material stability in their tested conditions. Two tested suggestions were 3,5-difluoropyridine-2-carboxylic acid and 5-hydroxy-2-methylbenzoic acid. This does not establish industrial operating lifetime or broad superiority over expert-selected additives.
Technology: Perovskite-R1, instruction-tuned QwQ-32B large language model
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The primary article reports the study-specific material and measurement summarized here.
- Still unknown: Independent replication and commercial deployment are not established.
- AI role: The domain-tuned LLM proposes precursor additives; researchers design, synthesize and test selected additives in devices and stability experiments.
Evidence sources: https://www.nature.com/articles/s43246-026-01099-9
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Primary publisher indexed full text/abstract and version metadata reviewed..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The AI-assisted EnzySFC workflow experimentally validated 10 previously uncharacterized nitrilases selected from 1,113 candidates. A wild-type Phototrophicales nitrilase showed exclusive nitrile hydratase activity; a separate Actinomycetia nitrilase was AI-evolved, with four tested top variants reaching 100% amide yield and complete substrate conversion.
Technology: EnzySFC AI-assisted enzyme discovery and functional-evolution platform
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The AI-assisted EnzySFC workflow experimentally validated 10 previously uncharacterized nitrilases selected from 1,113 candidates.
- Still unknown: The reported outcomes are enzyme and substrate assays; the abstract does not establish industrial scale performance or application in a living system. The 10 discovery candidates and 16 mutants were experimental follow-ups to model prioritization, not comprehensive validation of all 1,113 sequences.
- AI role: EnzySFC combined de novo enzyme discovery and multiobjective activity/specificity optimization to rank candidate nitrilases and propose evolved mutants. The study experimentally tested ten discovery candidates and sixteen mutants from one evolution prediction cycle; the named activity switch and product-selective variants were measured in assays.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/41645784/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A machine-learning-guided analysis of DESI multi-epoch spectra used UMAP to identify three systems with strongly varying spectral signatures that were proposed as new ‘double-faced’ white dwarfs with inhomogeneous surface composition. Follow-up phase-resolved time-series spectra verified the interpretation for one system; the other two remain candidates. This is one paper-level result and only the verified system is counted as confirmed.
Technology: UMAP dimensionality reduction applied to multi-epoch white-dwarf spectroscopy, alongside a neural-network spectral classifier
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: A machine-learning-guided analysis of DESI multi-epoch spectra used UMAP to identify three systems with strongly varying spectral signatures that were proposed as new ‘double-faced’ white dwarfs with inhomogeneous surface composition.
- Still unknown: The paper describes three new candidates, but only one has validating phase-resolved time-series spectra; the remaining two have limited/unequal signal-to-noise and are not included as confirmed objects. Existing classification of the first source as a DA+DB binary was challenged, not assumed disproven without the authors’ supporting phase-dependent evidence.
- AI role: Mapped spectral/photometric features and flagged white dwarfs with strongly changing multi-epoch spectra for expert review; this identified the systems subsequently investigated with phase-resolved spectroscopy.
Evidence sources: https://academic.oup.com/mnras/article/546/4/stag246/8466315
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Axiom Prover generated a proof of Fel's conjecture on syzygies of numerical semigroups, verified in Lean 4 with Mathlib, without hints, lemmas, or manual corrections after the run began. This is a case study and an unreviewed preprint, with no data on attempts or time taken.
Technology: Axiom Prover, Lean
Publication: Preprint · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: prospective or independent check
- Tested: The proof of Fel's conjecture was generated by AxiomProver with no hints, lemmas or manual corrections after kickoff and checked by Lean 4 with Mathlib; the formalization is publicly available and the mathematical results appear in a human-written arXiv preprint.
- Still unknown: The authors state the limiting factor is often the quality of the shared library (Mathlib), and the post describes a single case study with no claim about automating research.
- AI role: AxiomProver translated the statement into Lean, generated intermediate lemmas and assembled the complete Lean proof.
Evidence sources: https://axiommath.ai/research/proof-of-concept · https://arxiv.org/abs/2602.03716
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 4/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A cycle between GPT-5 and a robotic lab explored over 36,000 compositions and reduced protein-production cost by 40% compared with the previous baseline. This is an experimental result in one CFPS setup; it does not establish the same reduction for other proteins, scales, or facilities.
Technology: GPT-5
Family GPT-5: profile · models page
Publication: Announcement only · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: direct test
- Tested: Over six rounds in a cloud laboratory, the system tested more than 36,000 reaction compositions on 580 plates and reached a 40% lower protein production cost (57% lower reagent cost).
- Still unknown: Not stated in the reviewed material
- AI role: GPT-5 proposed and analyzed experiments in a closed loop with Ginkgo Bioworks automation.
Evidence sources: https://openai.com/index/gpt-5-lowers-protein-synthesis-cost/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A knowledge-augmented reinforcement-learning search over composition space, guided by learned property models and then validated by synthesis, identified a Zr-based bulk metallic glass, Zr63Cu15Al10Ni10Fe2. Single-step suction-cast specimens reached 1.7 GPa yield strength and 10.2% compressive plastic strain at room temperature. Across tested Fe-alloyed compositions, measured strength trends tracked model predictions; tungsten substitutions did not show the same ductility response.
Technology: AIMATDESIGN knowledge-augmented reinforcement-learning framework, using machine-learning property predictors and large-language-model-assisted model refinement and domain-knowledge rewards
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: A knowledge-augmented reinforcement-learning search over composition space, guided by learned property models and then validated by synthesis, identified a Zr-based bulk metallic glass, Zr63Cu15Al10Ni10Fe2.
- Still unknown: The reported measurements are room-temperature compression tests on single-step suction-cast samples; the paper notes that plastic strain was systematically lower than predicted for most tested specimens, partly because surface defects and residual stresses were not represented in the model. The 10.2% value is for the specific Fe2 composition and test protocol, not a general alloy-family property. This is distinct from a separate study of an ML-designed bulk metallic glass optimized for hardness (DOI 10.1038/s41467-026-73008-0): the present result concerns Zr63Cu15Al10Ni10Fe2 and its strength/ductility balance.
- AI role: An RL agent explored high-dimensional alloy compositions using ML classification/regression predictions as rewards and feasibility guidance. LLM-assisted feedback corrected unreliable model predictions and expert materials rules informed late-stage rewards. The authors used this workflow to choose Zr-based bulk-glass compositions for synthesis and mechanical testing; the experimentally tested Zr63Cu15Al10Ni10Fe2 composition achieved the reported strength and ductility result. This AI-guided composition search is the discovery contribution; the LLM was one component of a larger RL/ML workflow.
Evidence sources: https://www.nature.com/articles/s41524-025-01894-x
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A self-attention neural-network wavefunction (MagNet), trained only by minimising the microscopic Hamiltonian of interacting electrons in a magnetic field, finds both topological liquid and electron-crystal ground states across a range of Landau-level mixing. The authors present it as a method that discovers correlated phases without assumptions. This is an arXiv preprint with computational results only, not yet checked experimentally or peer reviewed.
Technology: MagNet: self-attention neural-network variational wavefunction
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A self-attention neural-network wavefunction trained only by minimising the microscopic Hamiltonian found topological liquid and electron-crystal ground states across Landau-level mixing; the abstract reports computational results only.
- Still unknown: The page states no limitations and mentions no experimental checks; the arXiv page lists a PRX Intelligence journal reference, while the source record describes it as a preprint.
- AI role: The MagNet self-attention neural network served as the variational wavefunction, trained by energy minimisation without labelled data.
Evidence sources: https://arxiv.org/abs/2602.03927
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
An XGBoost model trained on measurements from 70 Fe–Co–Cr–Mn–Cu compositions selected a layered double hydroxide for oxygen evolution. The reported composition had a measured overpotential of 270 mV at 10 mA cm⁻², versus a model prediction of 261.3 mV; the paper also reports a 72-hour stability test under its protocol. This comparison does not establish universal catalyst superiority or full-cell performance.
Technology: XGBoost regression
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Publisher supporting information and indexed abstract identify the optimal Fe0.15Co0.10Cr0.30Mn0.30Cu0.15 composition, experimentally measured 270 mV overpotential, and 72 h operation.
- Still unknown: Independent replication and broader application are not established by this bounded review.
- AI role: XGBoost trained on 70 experimental compositions guides selection from a much larger composition space; researchers synthesize and characterize the selected catalyst.
Evidence sources: https://doi.org/10.1021/acscatal.5c07303.s001
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 16 Nov 2026. Review scope: Publisher indexed article abstract and publisher-hosted supporting information; source identity/DOI and results checked. Full article text not accessed..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Researchers at the Zelinsky Institute used unsupervised machine learning to filter and cluster candidate atom-economic cycloaddition reactions generated from quantum-chemical data, then used a digital co-expert and chemist review to prioritize experiments. The team experimentally confirmed two previously undescribed silver-catalyzed cycloaddition reactions yielding new products within about one week.
Technology: Machine-learning candidate filtering and clustering combined with the digital co-expert reaction-prioritization workflow
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Researchers at the Zelinsky Institute used unsupervised machine learning to filter and cluster candidate atom-economic cycloaddition reactions generated from quantum-chemical data, then used a digital co-expert and chemist review to prioritize experiments.
- Still unknown: Two reactions were confirmed from a heavily filtered set of computational candidates, with final selection and validation by chemists. The reported approximately 180-fold acceleration is a comparison for the expert-screening bottleneck, not a general chemical-discovery speedup. The abstract does not specify individual products or reaction yields; no such details are inferred here.
- AI role: Unsupervised ML filtered and clustered computationally generated reaction candidates to reduce the set for review; the digital co-expert helped prioritize candidates. Human chemists made the final selection and performed experimental validation. The source reports two newly confirmed cycloadditions; the system was a human-AI pipeline, not autonomous discovery.
Evidence sources: https://onlinelibrary.wiley.com/doi/10.1002/ange.202523905
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
An unsupervised model trained on nearly 2,000 fossil footprints and millions of simulated variations identifies eight features of variation and agrees with experts about 90% of the time. Some tracks over 200 million years old resemble bird feet, which could suggest bird origins tens of millions of years earlier. The interpretation is open: they could be early dinosaurs whose feet happened to look similar.
Technology: DinoTracker (unsupervised model)
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: An unsupervised variational autoencoder was trained on 1,974 fossil footprints; after labeling, its groupings agreed with published expert identifications in 80-93% of cases (the press release says about 90%).
- Still unknown: The interpretation is open: the tracks could be early dinosaurs whose feet resembled birds' feet; the study states no limitations section.
- AI role: An unsupervised disentangled variational autoencoder found eight shape-variation features without prior labels, and predicted which dinosaur made each track.
Evidence sources: https://www.ed.ac.uk/news/ai-sheds-light-on-mysterious-dinosaur-footprints · https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=DOI:10.1073/pnas.2527222122&resultType=core&format=json
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The neural network scanned about 100 million cutouts from the Hubble Legacy Archive in roughly two and a half days: nearly 1,400 anomalies were flagged and over 1,300 confirmed, including more than 800 not previously documented in the literature (mergers, lenses, and jellyfish galaxies). AI only ranks candidates; the two counts are not explained, and no false-positive rate is given.
Technology: AnomalyMatch (neural network)
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: The neural network screened about 100 million Hubble Legacy Archive cutouts, and the authors personally inspected the top-ranked sources; more than 1,300 were confirmed as true anomalies, of which more than 800 were not previously documented; no false-positive rate is given.
- Still unknown: The release gives no false-positive rate, does not state how many candidates were inspected, notes that several dozen objects defied classification, and says expert review is still required.
- AI role: AnomalyMatch, a neural network, screened the archive and ranked candidate anomalies for expert inspection.
Evidence sources: https://esahubble.org/news/heic2603/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: press release. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Citations: at percentile 73 of works of the same type, year and subfield (OpenAlex work metrics, 5 citations; as of 15 Nov 2026). Attention, not a measure of correctness; not part of the score.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
An AI-accelerated workflow screened more than 1.3 million material structures, combining a graph-neural-network model of superconducting properties with density-functional calculations and stability filters. The researchers experimentally synthesized and confirmed superconductivity in Be₂HfNb₂ and Be₂Hf₂Nb, reporting two previously unreported compounds among 741 computationally screened stable candidates. The two compounds were confirmed in the Be₂Nb₃ substitution series; other listed candidates remain computational predictions.
Technology: BEE-NET equivariant graph neural networks in an AI-accelerated high-throughput screening workflow
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: An AI-accelerated workflow screened more than 1.3 million material structures, combining a graph-neural-network model of superconducting properties with density-functional calculations and stability filters.
- Still unknown: The AI model accelerated candidate selection but did not itself perform the synthesis; DFT and laboratory confirmation were essential. The paper reports an overall screening precision of 86%, so most candidate predictions were not experimentally verified in this study.
- AI role: Predicted Eliashberg spectral functions and superconducting transition temperatures to prioritize candidate structures, followed by DFT stability calculations and experimental synthesis/characterization by the research team.
Evidence sources: https://www.nature.com/articles/s41524-026-01964-8
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A machine-learning phase-picking and template-matching workflow expanded the catalogue for the 2018 Sierra Negra eruption onset, identifying a northwestern-caldera swarm of long-period microearthquakes missing from earlier manually picked records. The authors report the swarm began about two hours after a Mw 5.4 intra-caldera earthquake and preceded edifice failure and magma intrusion. The proposed fluid-triggering sequence is an interpretation of retrospective seismic and geodetic observations, not a prospective eruption forecast.
Technology: Volpick EQTransformer-based seismic phase picker combined with template matching and geodetic analysis
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: An EQTransformer-based phase picker and template matching expanded the seismic catalogue for the 2018 Sierra Negra eruption onset. The authors report 1,224 volcano-tectonic and 711 long-period events and identify a long-period swarm missing from the earlier manually picked catalogue, beginning about two hours after the Mw 5.4 earthquake.
- Still unknown: The proposed fluid-triggering sequence is an interpretation of retrospective seismic and geodetic observations; this study does not establish prospective eruption prediction.
- AI role: The learned EQTransformer-based phase picker supported a denser event catalogue; combined with template matching, this exposed a previously unreported long-period swarm in the archived seismic record.
Evidence sources: https://www.nature.com/articles/s41467-026-68645-4
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: Primary Nature Communications indexed article text and PDF-indexed article content, including abstract, Results, Methods, metadata and author-affiliation section..
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A human-LLM pipeline (Co-FunSearch) finds adversarial instances for heuristics, raising the lower bound for Best-Fit bin packing in the random-order model from 1.3 to 1.5. The authors prove the 3/2 ratio analytically and report a search-found instance scoring 1.497 over 10,000 random shuffles against Packingsolver optima. Caveat: arXiv preprint; human analysis supplies the final proofs.
Technology: Co-FunSearch (FunSearch-based LLM program search)
Family FunSearch: profile · models page
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: retrospective or benchmark
- Tested: Human-refined FunSearch outputs were used to derive improved adversarial lower bounds for standard heuristics; the page gives no details of the bin-packing test setup.
- Still unknown: Not reported on the abstract page; the full paper is needed for the bin-packing evaluation details.
- AI role: LLM-based evolutionary search (FunSearch) supplied initial patterns, which human experts refined and extrapolated into constructions.
Evidence sources: https://arxiv.org/abs/2601.16849
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A literature text-mining pipeline extracted phosphor/emission examples for model development; four Eu2+-activated candidates were synthesized and characterized. In the author manuscript, CaGd2S4:Eu2+ emitted at 440 nm versus 451 nm predicted under 365 nm excitation. Formula Li2SiN2 in one abstract conflicts with Li2SiN3 in results; this record does not claim that candidate formula.
Technology: Literature text-mining pipeline; genetic-algorithm feature selection; ML wavelength regression
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Author manuscript reports synthesized and measured CaGd2S4:Eu2+ and three other candidates; specific comparison retained.
- Still unknown: One formula conflict is unresolved; author manuscript/date/version details and composition-level prior art not fully checked.
- AI role: Text mining builds experimental training set; ML predicts emission wavelength and guides candidate selection.
Evidence sources: https://www.nature.com/articles/s41524-026-01967-5
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Primary source checked for identity, date, model role and bounded reported result..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
PolyCLOVER integrated multi-stage learning, active learning and high-throughput experiments across a ~100,000-member poly(β-amino ester) space, identifying three lead nanoparticles with measured antibacterial activity and in-vivo mouse outcomes. Mouse groups were small; this does not establish human efficacy or safety.
Technology: Multi-stage self-supervised learning and active learning coupled to high-throughput experimentation
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Primary Nature abstract/full text reports experimental MICs of 4 and 8 μg/mL against named resistant bacteria and mouse infection models; this is one study, not three discoveries.
- Still unknown: Independent replication and exhaustive prior-art review are not established.
- AI role: Multi-stage self-supervised learning and active learning coupled to high-throughput experimentation
Evidence sources: https://www.nature.com/articles/s41467-026-68682-z
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Primary source checked for study identity, stated AI contribution, date and bounded result..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Anthropic describes three cases: Biomni (Stanford) completed a GWAS analysis in 20 minutes in an initial trial and suggested new transcription factors in embryonic development; MozzareLLM (MIT and Whitehead) interprets gene clusters from CRISPR screens; a third lab is still testing. This is an Anthropic account without independent evaluation; some figures are estimates.
Technology: Claude (Biomni, MozzareLLM)
Family Claude: profile · models page
Publication: Announcement only · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Announcement only
- Test type: retrospective or benchmark
- Tested: Biomni's case studies (GWAS analysis, cloning design blind-evaluated against a postdoc, embryonic-tissue cell analysis) were checked by the Biomni team, citing a bioRxiv preprint; MozzareLLM was assessed qualitatively by its lab, with no quantitative benchmark.
- Still unknown: The article says the validation is not independent, some figures are estimates (such as the three-week human comparison), and Biomni cannot yet do everything out of the box.
- AI role: A Claude-powered agent (Biomni) analysed GWAS and embryonic cell data and suggested transcription factors; a Claude-powered system (MozzareLLM) interprets gene clusters from CRISPR screens.
Evidence sources: https://www.anthropic.com/news/accelerating-scientific-research
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: blog. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
DeepMet, an LSTM chemical language model trained on known human metabolites, generated a ranked space of metabolite-like structures and was integrated with mass spectrometry data. Researchers report identifying several dozen previously unrecognized mammalian metabolites across standards, human biofluids, mouse tissues and repository data. Some standard-supported assignments still have isomer ambiguity; the study also found known metabolites missing or misannotated in HMDB.
Technology: DeepMet chemical language model (three-layer LSTM over SMILES), integrated with MS/MS-spectrum prediction and meta-learning
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Researchers report identifying several dozen previously unrecognized mammalian metabolites across standards, human biofluids, mouse tissues and repository data.
- Still unknown: A subset of assigned structures retain regioisomer/stereoisomer ambiguity; human oversight helped prioritize structures for synthesis. The model explores chemical space similar to its training set and may miss divergent biosynthetic chemistry. The paper distinguishes novel metabolites from known compounds missing in HMDB.
- AI role: Learned metabolite-like chemical structures and prioritized structures likely to exist; combined with accurate mass/MS/MS and reference standards to support identification of previously unrecognized metabolites.
Evidence sources: https://www.nature.com/articles/s41586-025-09969-x.pdf
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Using a high-throughput qPCR assay, Arizona State University researchers measured microRNA 5p/3p strand use across six Caenorhabditis elegans developmental stages and selected tissues. A machine-learning model trained on experimentally measured strand usage identified compositional and structural sequence features associated with strand preference; the learned relationships predicted strand usage in other vertebrates, including humans, and the authors report that the rules are conserved but context-dependent.
Technology: Predictive machine-learning model using 77 biologically informed features, trained on high-throughput qPCR measurements from the HiTmiSS platform
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Using a high-throughput qPCR assay, Arizona State University researchers measured microRNA 5p/3p strand use across six Caenorhabditis elegans developmental stages and selected tissues.
- Still unknown: The dataset measured strand usage in C. elegans across six developmental stages and specific tissues; cross-species including human results are model predictions/validation against available strand-usage data, not direct tests of gene-regulatory outcomes in human cells. The result concerns strand-selection patterns and predictors, not disease causality or therapeutic efficacy.
- AI role: The model learned strand-preference relationships from the experimental strand-usage dataset, quantified feature contributions, and predicted preference across additional nematode microRNAs and vertebrate species. Experimental qPCR measurements established the C. elegans profiles; predictions across species extend the analysis but are not themselves experimental confirmation of function in each organism.
Evidence sources: https://academic.oup.com/nar/article/54/2/gkaf1510/8425383
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The self-iterative hierarchical ensemble model (SIM) nominated six high-confidence prokaryotic Argonaute candidates; five were experimentally validated as RNA-targeting proteins. The study identified three uncharacterized Agos with a novel N-terminal domain, defining a previously unrecognized subclass. Biochemical and in-vivo experiments on Haloferax profundi Ago (HpAgo) established RNA cleavage and a distinctive RNA-modification-sensing capability, which the authors used to build a method for quantifying modified RNA.
Technology: Self-Iterative hierarchical ensemble Model (SIM), an ensemble-learning and self-training pipeline
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The self-iterative hierarchical ensemble model (SIM) nominated six high-confidence prokaryotic Argonaute candidates; five were experimentally validated as RNA-targeting proteins.
- Still unknown: Five of six selected candidates were validated; the broader predicted sequence space was not experimentally tested. The specific RNA modification-sensing mechanism and assay performance are supported for HpAgo and tested conditions, not all Argonautes. No therapeutic or universal detection claim is made. Primary abstract/citation record checked; no independent replication assessed.
- AI role: SIM iteratively expanded its training labels to rank prokaryotic Argonautes for RNA targeting from sparse initial data; it prioritized candidates and identified a previously unrecognized domain-containing subclass. The researchers validated RNA cleavage and RNA-modification sensing biochemically and in vivo.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/42039271/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
SurvFinder analyzed 6,950 H&E slides from 1,604 stage II colorectal cancer patients across four cohorts and identified tertiary lymphoid structures (TLSs) as prognostic tissue features. Its multi-view model integrated TLS morphology and spatial location; explainable analysis highlighted TLS maturity and position near the tumor periphery as features associated with relapse-free survival and adjuvant-chemotherapy outcome patterns across the retrospective datasets.
Technology: SurvFinder, a multi-component deep-learning pipeline: WSINet for prognosis-associated tissue features, SegNet for TLS segmentation/subtyping, and MVNet for multi-view risk prediction
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: SurvFinder analyzed 6,950 H&E slides from 1,604 stage II colorectal cancer patients across four cohorts and identified tertiary lymphoid structures (TLSs) as prognostic tissue features.
- Still unknown: This was a retrospective analysis of patients enrolled in 2012–2018; results do not establish prospective clinical utility. Associations between predicted risk, TLS features and adjuvant-chemotherapy outcomes are observational and do not prove treatment benefit or that TLS features causally determine relapse. The authors identify absence of prospective validation and real-world deployment as limitations.
- AI role: Deep learning autonomously identified prognosis-associated tissue features in whole-slide images, segmented and subtyped TLSs, and integrated morphological and spatial features for risk prediction. Explainable-AI attribution identified TLS location near the tumor margin and TLS maturity state as important prognostic features.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/41532433/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
GPT-5.2 Pro produced the argument and Harmonic's Aristotle a Lean proof of infinitely many triples with a logarithmic gap in a factorial divisibility condition. The arXiv writeup calls it resolved autonomously, which is the author's claim; no independent review is stated.
Technology: GPT-5.2 Pro and Aristotle (Lean)
Family GPT-5: profile · models page
Publication: Preprint · Test context: prospective or independent check. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: prospective or independent check
- Tested: The argument produced by GPT-5.2 Pro was formalized as a Lean proof by Harmonic's Aristotle; the proof is machine-checked, and the abstract describes no independent human review.
- Still unknown: The abstract states no limitations and describes only the formal Lean proof as verification.
- AI role: GPT-5.2 Pro produced the argument; Aristotle produced the Lean proof, which the author translated into informal mathematics.
Evidence sources: https://arxiv.org/abs/2601.07421
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 4/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The study experimentally tested the previously model-predicted drug tranylcypromine against L. major promastigotes and intracellular amastigotes. It reports IC50 values of 83.6 and 31.6 µg/mL, respectively, and macrophage CC50 of 282.3 µg/mL. These relatively high in-vitro concentrations do not establish achievable exposure, patient safety or clinical efficacy; human monoamine oxidase mechanism is not shown as the parasite target.
Technology: Prior machine-learning drug-repurposing pipeline (the present study is experimental validation)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Primary full text reports promastigote and intracellular amastigote assays with dose responses and host-cell cytotoxicity.
- Still unknown: Prior drug/species work, tolerability and mechanism remain unresolved; no clinical inference.
- AI role: Study experimentally validates a model-prioritized repurposing candidate from earlier research.
Evidence sources: https://doi.org/10.3389/fmicb.2025.1756231
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: publisher primary full text; PubMed primary metadata/abstract. Material checked for the specific claims stated in the record..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The January 2026 Science abstract reports DrugCLIP screening roughly 10,000 human proteins against 500 million compounds. Experimental hits included 15% for the norepinephrine transporter, with structures of two inhibitor complexes, and 17.5% for TRIP12 using AlphaFold2 structures. Most database results remain computational. The work also has a bioRxiv preprint first indexed in September 2024.
Technology: DrugCLIP (deep contrastive learning)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: Biochemical hit testing and structures of two NET inhibitor complexes.
- Still unknown: Not assessed: full assay denominators, uncertainty and genome-wide experimental coverage.
- AI role: Contrastive learning ranked protein-compound pairs.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/41505557/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 4/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A random-forest regression model trained on seismic waveform features predicted GNSS-derived displacement-rate changes over roughly six years in western Shikoku, Japan. Comparing detections with earlier catalogs, the authors report five previously unreported short-term slow-slip events among 23 detected episodes; GNSS deformation patterns and fitted fault models support interpreting the transients as slow slip. The paper-level result is a retrospective detection method and event set, not prospective earthquake prediction.
Technology: Random-forest regression model
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A random-forest regression model trained on seismic waveform features predicted GNSS-derived displacement-rate changes over roughly six years in western Shikoku, Japan.
- Still unknown: The study is retrospective and geographically limited to western Shikoku. Model-predicted rates can differ from observations; smoothing can make events appear longer, a practical threshold affects counts (raising it leaves two newly identified events), and the authors state that events without associated tremor may be missed. It does not establish a prospective earthquake-warning capability or a causal trigger for the event that followed the 2016 Kumamoto earthquake.
- AI role: Mapped statistical features from continuous seismic waveforms to GNSS-derived displacement rates and flagged sharp predicted increases as possible short-term slow-slip events; authors then fitted fault models and compared deformation with existing event catalogs.
Evidence sources: https://link.springer.com/article/10.1186/s40623-025-02331-5
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
CleaveNet combines a generative model conditioned on protease cleavage profiles with a predictor that ranks candidate sequences. In an in-vitro screen of 95 peptide substrates against 12 matrix metalloproteinases, the authors report that designed substrates were cleaved by MMP13 and identify substrates with different efficiency/selectivity trade-offs. The panel includes training controls and baseline candidates; it is not 95 novel AI discoveries. These are laboratory substrates, not validated diagnostic or therapeutic products.
Technology: CleaveNet generator and predictor; deep learning
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: The paper reports the specific in-vitro protein assay summarized here.
- Still unknown: Clinical or diagnostic deployment and independent replication are not established.
- AI role: A generator proposes peptide sequences conditioned on cleavage profiles; a model ranks candidates and researchers test synthetic substrates against recombinant enzymes.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC12775482/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 16 Nov 2026. Review scope: Full primary journal article text, results and experimental assays reviewed. The preprint version-specific claim correspondence was not checked..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
HEX, a pathology-image AI model, predicts spatial expression of 40 protein biomarkers from routine H&E slides. In retrospective NSCLC cohorts, combining HEX-derived virtual spatial proteomics with histology improved prognosis prediction and, in a separate 148-patient immune-checkpoint-inhibitor cohort, improved response prediction over H&E-only and virtual-proteomics-only models. The authors interpreted virtual maps as showing co-localized T-helper and cytotoxic T-cell niches among responders and immunosuppressive macrophage/neutrophil aggregates among non-responders.
Technology: HEX (H&E to protein expression), using a pathology foundation-model backbone; MICA (multimodal integration via co-attention)
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: In retrospective NSCLC cohorts, combining HEX-derived virtual spatial proteomics with histology improved prognosis prediction and, in a separate 148-patient immune-checkpoint-inhibitor cohort, improved response prediction over H&E-only and virtual-proteomics-only models.
- Still unknown: Retrospective observational cohorts; predictive associations do not establish causal mechanisms or improved patient outcomes from model-guided treatment. The 148-patient immunotherapy cohort is separate from the prognosis cohorts. The paper itself says extensive clinical validation remains necessary; do not describe the finding as a clinically validated diagnostic or prospective treatment-selection tool.
- AI role: HEX predicts expression of 40 proteins and produces spatially resolved virtual proteomics from H&E images; MICA fuses the original image with AI-derived maps for outcome prediction. Researchers used predicted marker distributions to interpret therapy-associated tumor–immune spatial patterns. These are retrospective model analyses with cohort-based clinical associations, not prospective treatment evidence.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC12823406/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A machine-learning pipeline selected a seven-gene lactylation-related transcriptomic score (LDS) from nine Alzheimer disease cohorts. The study evaluated the signature in seven independent brain transcriptomic datasets and a single-center clinical plasma cohort of 540 participants (180 controls, 90 with amnestic mild cognitive impairment and 270 with Alzheimer disease). Plasma LDS had AUC 0.772 for Alzheimer disease; combining LDS with p-tau181 and p-tau217 had AUC 0.859. LDS scores increased across control, aMCI and Alzheimer groups and correlated with MMSE and Braak stage.
Technology: Lactylation-Derived Score (LDS), a random-forest feature-screening and partial least-squares logistic-regression model
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: retrospective or benchmark
- Tested: A machine-learning pipeline selected a seven-gene lactylation-related transcriptomic score (LDS) from nine Alzheimer disease cohorts.
- Still unknown: The plasma validation cohort was recruited at one center and the analyses are retrospective and observational. The score does not establish lactylation as a causal mechanism, clinical benefit, or readiness for routine diagnosis; broader and prospective validation is needed. This is a transcriptomic signature and a different result from previously reported Alzheimer spectral-protein/chemical biomarker studies.
- AI role: Machine learning selected a seven-gene expression signature from lactylation-related genes across multi-cohort brain transcriptomic data; the resulting score was measured by RT-qPCR in a separate clinical plasma cohort and evaluated alone and with plasma p-tau markers.
Evidence sources: https://www.nature.com/articles/s41746-025-02305-7
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
A fine-tuned Uni-Mol model used simulation-derived data to screen 160,000 tetrapeptides. Researchers experimentally tested 26 selected tetrapeptide candidates: 15 formed β-sheet assemblies, including seven hydrogelators, as assessed with structural and assembly assays. The study reports experimental validation across 46 candidates overall; the 15/26 result is specific to the screened tetrapeptide set. This is a materials-science preprint, not evidence of medical function or clinical use. A related 2023 study explored tetrapeptide hydrogelation; this record makes no claim that every sequence is unprecedented and is limited to the present paper’s β-sheet structural validation.
Technology: Fine-tuned Uni-Mol machine-learning model for Amyloid-Like Tendency prediction, guided by hybrid-resolution molecular-dynamics simulations
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Preprint
- Test type: direct test
- Tested: In the Uni-Mol-screened tetrapeptide set, 26 candidates were experimentally characterized; 15 formed β-sheet assemblies and seven were reported as hydrogelators. The authors also report validation of 46 candidates across the study overall.
- Still unknown: The 46-candidate set combines tripeptide and tetrapeptide validations; the 15/26 result is specific to the tetrapeptides. The work is a preprint and establishes laboratory assembly outcomes only, not biomedical function, clinical utility, or independent replication. No claim is made that every sequence is unprecedented.
- AI role: Uni-Mol was fine-tuned on simulation-derived data and used to screen 160,000 tetrapeptides for Amyloid-Like Tendency; predictions guided candidate selection for further simulation and experimental structural characterization.
Evidence sources: https://www.cambridge.org/engage/chemrxiv/article-details/694d4b3be3cb457e13d19e29 · https://www.cambridge.org/engage/api-gateway/coe/assets/orp/resource/item/694d4b3be3cb457e13d19e29/original/discovery-of-sheet-peptide-assembly-codes-via-an-experimentally-validated-predictive-computational-platform.pdf · https://s3.eu-west-1.amazonaws.com/assets.prod.orp.cambridge.org/02/54654caa1745248e0618e2e4ca692a.pdf
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 17 Nov 2026. Review scope: ChemRxiv v1 page, full manuscript and Supplementary Information; official bioRxiv API metadata/abstract for cited 2025 computational preprint; primary 2023 hydrogel paper and linked supplementary-data index. Full-text source limits and the bounded result are stated..
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
The January 2026 Nature Biotechnology abstract reports 40 small-molecule hits across ten disease-associated RNA targets, with nanomolar-to-micromolar binding. One predicted compound reduced MYC expression, inhibited proliferation and promoted apoptosis in three cancer cell lines. SMRTnet combines language models with convolutional and graph attention networks, using RNA secondary structure. These are laboratory results.
Technology: SMRTnet (language-model-based)
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: RNA-binding assays and cancer-cell experiments.
- Still unknown: Not assessed: full-paper validation rates and therapeutic performance.
- AI role: Multimodal deep learning predicted small molecule-RNA interactions.
Evidence sources: https://pubmed.ncbi.nlm.nih.gov/41482542/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 4/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.
Standigm ASK’s knowledge-graph model ranked Parkinson’s disease target associations; subgraph enrichment and expression analyses narrowed 2,527 predicted targets to five candidates, including TPP1. In a human cell-based α-synuclein seeding model, TPP1 knockdown increased aggregate accumulation and insoluble α-synuclein, experimentally supporting a TPP1–α-synuclein relationship and a possible protective role.
Technology: Standigm ASK knowledge-graph path-based deep-learning model with subgraph-level over-representation re-ranking
Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.
Result, AI contribution and limits
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
- Publication status: Peer-reviewed paper
- Test type: direct test
- Tested: In a human cell-based α-synuclein seeding model, TPP1 knockdown increased aggregate accumulation and insoluble α-synuclein, experimentally supporting a TPP1–α-synuclein relationship and a possible protective role.
- Still unknown: The functional experiment is a human neuroblastoma cell model exposed to α-synuclein preformed fibrils, not a clinical or in-vivo therapeutic demonstration. Knockdown increasing aggregation supports a protective association; direct enzymatic cleavage of α-synuclein is proposed from structural modeling, not established by the reported work. TPP1 is a candidate therapeutic target, not a validated treatment.
- AI role: The model integrated heterogeneous biomedical databases, encoded relational paths using a convolutional network and attention aggregation, and produced disease-target scores. Researchers applied enrichment and transcriptomic filters to prioritize TPP1, then validated TPP1’s relationship to α-synuclein aggregation via siRNA knockdown and cellular assays; AlphaFold-Multimer modeling suggested, but did not prove, direct proteolysis.
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC12814083/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review scope and date are not recorded; source checked does not establish full-text review.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.
Follow-up and relationships
No follow-up recorded; this is not a claim that no follow-up exists.
Follow-up review scope and date are not recorded.
No source-supported relationships recorded.