GDM Self-Reasoning

Benchmark / dataset

Situational awareness / self-modification tasks.

Category: Deception, scheming & alignment auditing
Insurance exposure: Agent actions
Publisher: Google DeepMind (ported)
Publisher type: Frontier lab
License: MIT
Status: In inspect_evals (active)
Last commit: 2026-10-15
Stars: 704(shared repo star count — UKGovernmentBEIS/inspect_evals backs 19 entries in this catalog)

Wanich Risk Lab, 2026.