Gemini 2.5
Google DeepMind — Gemini
rich detail 12 method fields published
Evaluation
CBRN open-ended qualitative questions (domain-expert assessed)
Qualitative assessment for biological, radiological and nuclear domains: knowledge-based, adversarial and dual-use content spanning difficulty from non-expert to PhD-expert, and threat journeys; domain-expert led
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.