Measuring the Depth of LLM Unlearning via Activation Patching

2 points | by jaeunglee an hour ago

No comments yet