作者位次 #9/13ACL 2026Poster· Natural Language Processing
Mitigating Lost in Multi-turn Conversation via Curriculum RL with Verifiable Accuracy and Abstention Rewards
Ming Li · Pei Chen · Zhenhao Zhang · Tao Yang · Xinyang Zhang · Han Li · Tianyu Cao · Ming Zeng · Zhuofeng Wu · Meng Jiang · Huasheng Li · Lihong Li · Bing Yin
摘要
Large Language Models demonstrate strong capabilities in single-turn instruction following but suffer from Lost-in-Conversation (LiC), a degradation in performance as information is revealed progressively in multi-turn settings. Motivated by the current progress on Reinforcement Learning with Ver...