Article

BiomniBench: Process-level Evaluation of LLM Agents for Real-world Biomedical Research

2026-05-14

Abstract excerpt

LLM agents now perform real biomedical research, but evaluating them rigorously is hard. Outcome-only benchmarks fail in two ways. First, a correct final answer can come from memorization, reward hacking, or wrong reasoning that produces the right number by chance. Second, valid alternative analyses are marked wrong simply because they differ from the reference. We introduce BiomniBench, a process-level evaluation...

Topics

Open a Topic to create a Post that cites this publication.

Identifiers and source

Literature Corpus work: c682b60e-292c-5c57-8f9a-ebc91296b43d
DOI: 10.64898/2026.05.12.724604

Open publication

Related research

Semantic proximity does not establish scientific evidence.

Click a neighbor to travelStep 1 · 12 closest

BiomniBench: Process-level Evaluation of LLM Agents for Real-world Biomedical ResearchDOI 10.64898/2026.05.12.724604Read

BioML-bench: Evaluation of AI Agents for End-to-End Biomedical MLSemantic proximity Read

Agentic systems are adept at solving well-scoped, verifiable problems in computational biologySemantic proximity Read

GeneBench-Pro: Evaluating Multistage Statistical Reasoning in Genomics, Quantitative Biology, and Translational BiomedicineSemantic proximity Read

ClaroAI-Bench: Evaluating Agentic Scientific Reproducibility on Real Biomedical PapersSemantic proximity Read

Biomni: A General-Purpose Biomedical AI AgentSemantic proximity Read

GeneBench: Assessing AI Agents for Multi-Stage Inference Problems in Genomics and Quantitative BiologySemantic proximity Read

An LLM-assisted framework for accelerated and verifiable clinical hypothesis testing from electronic health recordsSemantic proximity Read

Evaluating open LLMs for agentic analysis orchestration in a typical biomedical labSemantic proximity Read

Agentomics: An Agentic System that Autonomously Develops Novel State-of-the-art Solutions for Biomedical Machine Learning TasksSemantic proximity Read

MedGenesis: Toward a World Model for Autonomous Clinical and Translational ResearchSemantic proximity Read

Multi-agent AI enables evidence-based cell annotation in single-cell transcriptomicsSemantic proximity Read

MedAgent: A Retrieval-Augmented Clinical Decision Support Agent with Verifiable Evidence Grounding for Evidence-Based MedicineSemantic proximity Read