Aligning Large Language Models with Representation Editing: A Control Perspective

Open in new window