Position: The Hidden Costs and Measurement Gaps of Reinforcement Learning with Verifiable Rewards 文章
ArXiv CS.AI2026-05-27NEWSen作者: Fang Wu, Aaron Tu, Weihao Xuan, Heli Qi, Xu Huang, Qingcheng Zeng, Shayan Talaei, Yijia Xiao, Peng Xia, Xiangru Tang, Yuchen Zhuang, Yinxi Li, Bing Hu, Hanqun Cao, Wenqi Shi, Rui Yang, Nan Liu, Huaxiu Yao, Ge Liu, Li Erran Li, Amin Saberi, Naoto Yokoya, Jure Leskovec, Yejin Choi
Position: The Hidden Costs and Measurement Gaps of Reinforcement Learning with Verifiable Rewards · 相关技术
相关技术
TTEODEPLALLMlanguage model递归自我改进远程代码执行（RCE）reinforcement learningevaldivide-and-conquer partitioningabstentionUCTTAMStraight-Through EstimatorStanScalaReinforcement learning with verifiable rewardsRLVRPromptParts-of-Speech (POS) tagsPROBENATlarge language modelsHuman-Robot InteractionGranular Alignment ParadigmForCamouflaged object detectionARGANN