作者位次 #6/12ACL 2026Poster· Natural Language Processing
T ri P lay- RL : Tri-Role Self-Play Reinforcement Learning for LLM Safety Alignment
Zhewen Tan · Wenhan Yu · Jianfeng Si · Tongxin Liu · Kaiqi Guan · Huiyan Jin · Jiawen Tao · Xiaokun Yuan · Xiangzheng Zhang · Duohe Ma · Tong Yang · Lin Sun
摘要
In recent years, safety risks associated with large language models have become increasingly prominent, highlighting the urgent need to mitigate the generation of toxic and harmful content. The mainstream paradigm for LLM safety alignment typically adopts a collaborative framework involving three...