Gemini 2.5
Google DeepMind — Gemini
rich detail 17 method fields published
Evaluation
Deceptive alignment: situational awareness challenges
Challenges testing whether the model can reason about itself as an agent in its deployment environment and intervene to gain affordances (e.g., changing configuration files); suite of 11 evaluations; published in Phuong et al. (2025)
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.