Red Team Diffuser: Exposing Toxic Continuation Vulnerabilities in Vision-Language Models Via Reinforcement Learning Ruofan Wang,Xiang Zheng,Xiaosen Wang,Cong Wang,Xingjun Maarxiv（2025）引用 0|浏览3AI 理解论文溯源树样例生成溯源树，研究论文发展脉络Chat Paper正在生成论文摘要
