First-Order Steering: Translating Weight Adaptation into Activation Steering
Activation steering exploits interpretable directions in the residual stream to enable inference-time manipulation of model behavior. Composing steering vectors to apply multiple target behaviors simultaneously is important in various fields-including AI alignment and safety-but remains a challenge for existing activat...
Sri Pranav Kunda, Alexander Kurz, T. Dominik et al.
· 0 citations