Back to papers
semanticscholar8.0 / 10

AI Alignment Evaluation Detection

Loni Hagen, A. Hagen, Daniel Tafmizi, Christopher Reddish, A. Fox, Lingyao Li, Nic DePaula

Abstract

Research on Claude Sonnet 3.7's ability to know when it's in alignment evaluations

Research area

agentic misalignmentagent misalignmentai alignment
Published
2025
Source
semanticscholar
Org
Apollo Research
View paper
Sign in to read and join the discussion.