Picture for Joe Needham

Joe Needham

Honesty to Subterfuge: In-Context Reinforcement Learning Can Make Honest Models Reward Hack

Add code
Oct 09, 2024
Viaarxiv icon