Evaluation
CBRN evaluations for Claude Opus 4 / Claude Sonnet 4 (RSP)
human uplift studies, expert red-teaming (bacterial and viral), knowledge/skill MCQs, open-ended pathway questions, task-based agentic evaluations with search and bioinformatics tools
Biological & chemical risks
Anthropic
How it was run
Open a release to see the setup, scoring, and other details that the lab published. Only fields the lab actually disclosed are shown.
Claude Opus 4, Claude Sonnet 4
Anthropic
rich detail
31 method fields published
See how this test was run