Caught in the Act: Probes Effectively Detect Sabotage and Catch Unverbalized Deception
By Oskar J. Hollinsworth · Paper · cs.LG
Recent incidents have highlighted the challenge of monitoring LLM agents and the danger of models deceiving people. We show that white-box deception detection via probes can be scaled up to frontier monitoring settings by collecting the largest deception dataset to date for train
Tags: Cs.lg