1 paper
Arthur Vogels, Benjamin Wong, Yann Choho +2
Activation steering methods control large language model (LLM) behavior by modifying internal activations at inference time. However, most existing activation steering methods rely…