AI Science Tracker
EN

Iowa State University

2 discoveries from 2022 to 2026

Discoveries

Agent-proposed vortex scale improves a reference-flow prediction

Source check: Source checked. Not independent scientific confirmation.

Finding: Claude-driven agents translated a human-supplied vortex-axis modeling idea into candidate subgrid length scales. The selected Liutex-based formulation predicted backward-step reattachment at 7.13 step heights versus the published experimental 6.67; the standard scale predicted 12.50. This is a selected computational closure result, not a new experimentally observed turbulence mechanism.

Technology: Claude Opus 5.5 discovery, review, implementation and evaluation agents

Family Claude: profile · models page

Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
LLM agents derived and implemented explicit vortex-axis-based formulas rather than merely executing a fully specified human closure. Subsequent numerical solvers evaluated their physical predictions.
Test and result
Candidate formulas were screened sequentially on four canonical/engineering flows using published references. The selected closure was similar to, or slightly better than, an expert shear-layer-adapted baseline.
Limitation / still unknown
The same cases drove selection, not an untouched validation suite. Five runs repeated proposal/review/implementation but did not repeat expensive flow simulations. Human concept, reference data and fallback assumptions constrain discovery. No universal closure superiority or real-aircraft benefit is established.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://arxiv.org/html/2610.29850v1

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: 21 Nov 2026. Review scope: Actual primary candidate-generation Methods, selection workflow, reference-flow Results/Table6 and physical fallback interpretation; current v1 date/status..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 1/5 · 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.

High-throughput image-based plant stand count estimation using convolutional neural networks

Source check: Source checked. Not independent scientific confirmation.

Finding / tested result: A convolutional network counted plant stands from field images and was evaluated against manual counts, demonstrating a specific crop-imaging measurement result.

Technology: Convolutional neural network

Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
Counts plant stands from field images and is evaluated against manual counts.
Limitation / still unknown
The reported result is limited to the study data, samples and conditions; broader application was not established.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://doi.org/10.1371/journal.pone.0268762 · https://europepmc.org/article/MED/35901120 · https://europepmc.org/articles/PMC9333310 · https://www.ncbi.nlm.nih.gov/research/bionlp/RESTful/pmcoa.cgi/BioC_json/PMC9333310/unicode

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: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

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.

Open the interactive profile for the timeline, or browse all organizations and model families.