Refuse without Refusal: A Structural Analysis of Safety-Tuning Responses for Reducing False Refusals in Language Models
Paper • 2609.04714 • Published • 6
mz-kim
mz-kim
·
AI & ML interests
None yet
Recent Activity
upvoted a paper 19 days ago
Knowing What Not to Answer: Selective Non-Compliance in Vision-Language Models new activity 29 days ago
huggingface/HuggingDiscussions:[FEEDBACK] Daily PapersOrganizations
upvoted 2
papers 19 days ago