AI BENCHMARK PROFILE

BixBench3

General AIKnowledge & Reasoning

Evaluates AI agents on 20 research-study-scale computational biology tasks that require processing raw data into scientific results, scored by matching generated artifacts against study-derived targets.

Released: 2026-08-26
Readiness: Paper only
Primary field: General AI

Why it matters

Systematically measures agent capability in long-horizon, data-intensive biological analyses, revealing gaps in sequential reasoning and resource management.

Motivation

Artificial intelligence (AI) promises to accelerate biological research by automating computational analyses.

Primary resources

Paper ↗Hugging Face ↗

Benchmark Radar records only publicly supported details and links back to primary sources for verification.