Author

Weiwei Hou

h-index 0 View on Semantic Scholar

We have 1 of 2 papers

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Preprint Aug 2026

When Skills Meet Safety: Benchmarking and Characterizing the Adaptive Jailbreak Robustness of Skill-Merged LLMs

SkillSafe-Bench is introduced, a controlled benchmark that scores skill-merged models on static refusal, adaptive jailbreak robustness, and capability retention under a conservative two-judge AND rule, and the static effect of merging is base-conditional.

Yu Ma, Hongli Shi, Jing Li et al. · 0 citations