MétaCan
Menu
Back to cohort
Record W4414028042 · doi:10.1101/2025.09.01.673319

BioML-bench: Evaluation of AI Agents for End-to-End Biomedical ML

2025· preprint· en· W4414028042 on OpenAlex

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

affAt least one author lists a Canadian institution in the pinned OpenAlex snapshot.

Bibliographic record

VenuebioRxiv (Cold Spring Harbor Laboratory) · 2025
Typepreprint
Languageen
FieldDecision Sciences
TopicScientific Computing and Data Management
Canadian institutionsUniversity of Toronto
Fundersnot available
KeywordsEnd-to-end principleBench to bedsideComputer scienceArtificial intelligenceMedicineMedical physics

Abstract

fetched live from OpenAlex

Abstract Large language model (LLM) agents hold promise for accelerating biomedical research and development (R&D). Several biomedical agents have recently been proposed, but their evaluation has largely been restricted to question answering (e.g., LAB-Bench) or narrow bioinformatics tasks. Presently, there remains a lack of benchmarks evaluating agent capability in multi-step data analysis workflows or in solving the machine learning (ML) challenges central to AI-driven therapeutics development, such as perturbation response modeling or drug toxicity prediction. We introduce BioML-bench , the first benchmarking suite for evaluating AI agents on end-to-end biomedical ML tasks. BioML-bench spans four domains (protein engineering, single-cell omics, biomedical imaging, and drug discovery) with tasks that require agents to parse a task description, build a pipeline, implement models, and submit predictions graded by established metrics (e.g., AUROC, Spearman). We evaluate four open-source agents: two biomedical specialists (STELLA, Biomni) and two generalists (AIDE, MLAgentBench). On average, agents underperform relative to human baselines, and biomedical specialization does not confer a consistent advantage. We also found that agents which employed more diverse ML strategies more often tended to score highest, suggesting that architecture and scaffolding may be stronger determinants of performance. These findings underscore both the potential and current limits of agentic systems for biomedical ML, and highlight the need for systematic, reproducible evaluations. BioML-bench is provided open-source at github.com/science-machine/biomlbench .

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.024
metaresearch head score (Gemma)0.012
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesMetaresearch, Meta-epidemiology (narrow)
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Not applicable · Consensus signal: Not applicable
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.315
Threshold uncertainty score1.000

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0240.012
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0010.000
Bibliometrics0.0020.003
Science and technology studies0.0000.000
Scholarly communication0.0010.000
Open science0.0030.003
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0010.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.130
GPT teacher head0.385
Teacher spread0.255 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it