Safety for Whom? Boundary-aware self-distillation for controlled LLM safety refusal
Read the original at huggingface.co→Original headline: "Safety for Whom? Refusing the Right Subset of a Topic, Not the Whole Topic"
Coverage timeline
- Sep 7, 04:00 UTC arXiv cs.CL Safety for Whom? Boundary-Aware Self-Distillation for Controlled LLM Safety Refusal
- Sep 8, 14:23 UTC Hugging Face lead source Safety for Whom? Refusing the Right Subset of a Topic, Not the Whole Topic