Google DeepMind releases Gram, an alignment auditing tool to detect AI agents secretly sabotaging tasks
Gram found a 3% natural sabotage rate in Gemini models
#213Neel Nanda@NEELNANDA5Auditing agents are a key way to catch misaligned models, but they need to audit realistically - any model will do bad things if pushed hard enough. I'm glad GDM is pushing on how to improve realism here!
David Lindner@davlindnerWill your AI agent secretly sabotage your work? Existing alignment evals don't directly answer this question Meet Gram: the alignment auditing tool we use to assess how likely AI agents are to engage in sabotage during internal deployments at @GoogleDeepMind
#1469Zac Kenton@ZACKENTON1Alignment auditing and evals, with an emphasis on simulating realistic (rather than red-team) settings. Interestingly, natural sabotage rates for Gemini (3%) are largely due to overeagerness: excessive roleplaying; too literal interpretation of instructions to optimise a goal
David Lindner@davlindnerWill your AI agent secretly sabotage your work? Existing alignment evals don't directly answer this question Meet Gram: the alignment auditing tool we use to assess how likely AI agents are to engage in sabotage during internal deployments at @GoogleDeepMind