作者位次 #3/4ICML 2026Poster· Deep Learning / Large Language Models
SPA: A Simple but Tough-to-Beat Baseline for Knowledge Injection
Kexian Tang · Jiani Wang · Shaowen Wang · Kaifeng Lyu
检索与知识OpenReview
首次获取该候选人的完整档案及 StartLux 评估共 20 积分；每位用户单独解锁，之后查看免费。主动刷新 20 积分。
实时查找电子邮箱，费用不超过20积分；未找到邮箱不扣费。
Kexian Tang · Jiani Wang · Shaowen Wang · Kaifeng Lyu
Shaowen Wang · Yuke Zheng · Tansheng Zhu · Shuang Chen · Shaofan Liu · Suncong Zheng · Li Jian
Shaofan Liu · Guoqiang Zhang · Shihan Dou · Huiyuan Zheng · Yiming Zhou · Junjie Ye · Shaowen Wang · Shichun Liu · Jiazheng Zhang · Tao Gui · Qi Zhang · Xuanjing Huang
Reward models (RMs) are the surrogate objectives in reinforcement learning from human feedback (RLHF), and their scores directly steer policy optimization. We show that standard RM training is vulnerable in data subsets where response quality depends only weakly on the context: such instances enc...