B3 (Bacterial Biothreat Benchmark): leaderboard

Pilot implementation of the Bacterial Biothreat Benchmark dataset from the Biothreat Benchmark Generation framework, assessing LLM biosecurity risk across bacterial threat tasks and actor-capability levels.

Source: arxiv.org.

Interactive version: theaggregate.ai/benchmark?slug=b3-bacterial-biothreat-benchmark · How It Works · Data refreshed daily, snapshot 2026-09-05.