SafeSteer: Localized On-Policy Distillation for Efficient Safety Alignment 文章

ArXiv CS.CL2026-06-02NEWSen作者: Hao Li, Jingkun An, Zijun Song, Pengyu Zhu, Rui Li, Hao Wang, Wendi Feng, Yesheng Liu, Lijun Li, Jin-Ge Yao, Lei Sha