SafetyAnalyst: Interpretable, Transparent, and Steerable Safety Moderation for AI Behavior

Devs

SafetyAnalyst: Interpretable, Transparent, and Steerable Safety Moderation for AI Behavior | Read Paper on Bytez