AI Science Tracker
EN

Allen Institute for Artificial Intelligence

1 discovery from 2026

Discoveries

LLM-assisted hypothesis generation identified an immune signal in invasive lobular breast cancer

Source check: Source checked. Not independent scientific confirmation.

Finding: An LLM hypothesis-generation workflow, with oncologist input, surfaced an immune-signature hypothesis in invasive lobular carcinoma. The authors assessed it in METABRIC and tissue analyses. This is a human–AI collaborative result from a non-peer-reviewed preprint and does not establish treatment benefit.

Technology: AutoDiscovery used large language models to generate and rank data-driven hypotheses; oncologist review guided follow-up.

Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.

AI contribution
AutoDiscovery used large language models to generate and rank hypotheses; oncologist review guided follow-up.
Test and result
The hypothesis was assessed in an independent METABRIC cohort and by multiplex immunofluorescence; oncologist selection contributed to the workflow.
Limitation / still unknown
This is a non-peer-reviewed preprint; findings do not establish response to immunotherapy or clinical benefit.
Evidence sources and annotation context

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

Evidence sources: https://www.medrxiv.org/content/10.64898/2026.08.25.26361365v1.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: 22 Nov 2026. Review scope: Primary full-text review as recorded in case evidence; record-level claim and attribution review..

No follow-up recorded; 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 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.

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