Reward Model Evaluation

Ask about this topic

Overview

Reward model evaluation studies whether learned judges match human preferences and provide useful signals for selecting or improving language-model and agent behavior. Researchers use pairwise and listwise preference benchmarks, calibration and uncertainty metrics, and stress tests across settings such as long-form writing, math, code, personalized dialogue, and robot or computer-use trajectories. Evaluations also test downstream effects in best-of-N selection and reinforcement learning, alongside methods such as human-verified data, LLM judges, and perturbation-based audits. A central tension is that strong benchmark accuracy may not transfer across domains or predict safe policy improvements, especially when long-horizon errors, surface biases, and reward hacking are involved.

Latest papers

CardsList

46 papers9 papers in the last four weeks, up 80% on the four weeks before. 0.1% of all new papers.

  1. AndroidFlux: Evaluating and Reward Modeling Failure Recovery Capabilities in Mobile Use Multimodal Agents

    Oct 25, 2026Gyubeum Lim, Yemo Koo, Ghunwoo Jang +1Multimodal Agent EvaluationReward Model Evaluation

    Georgia Institute of Technology · Konkuk University

  2. ACWM-RM: A Reference-Free Reward Model from Complementary Mismatch Preferences

    Oct 23, 2026Shibo Hong, Siyu Ren, FengJiao Chen +3Multimodal Reward ModelingRL for Video Generation

    College of Computer Science and Artificial Intelligence, Fudan University · Meituan

  3. What Pairwise Evaluation Misses: Reward Scores across Prompts

    Oct 22, 2026Mateusz Modrzejewski, Tomasz RadzikowskiReward Model EvaluationPairwise Preference Evaluation

    Warsaw University of Technology

  4. Scaling Laws for Generative Reward Models

    Oct 22, 2026Angel Raychev, Nathan Lile, Rafael RafailovReward ModelingRL for Language Models

    Stanford University · Thinking Machines

  5. LegalRewardBench: Benchmarking Reward Models for Legal Preference Evaluation

    Oct 13, 2026Hui Huang, Jie Zhou, Hongli Zhou +7Reward Model EvaluationPairwise Preference Evaluation

    Harbin Institute of Technology · ModelBest Inc. · The University of Osaka

  6. Benchmarking and Training Reward Models for Cross-Source Evidence Grounding in Search Agents

    Oct 13, 2026Haochen Zhao, Minsheng Hao, Zheng Li +8Evidence-Grounded GenerationAI Agent Benchmarks

    Peking University · Tencent · Renmin University of China

  7. UXBench Pro: Benchmarking Personalized User Experience in Multi-Turn Dialogue Interactions

    Oct 8, 2026Mengze Hong, Zeyang Lei, Wenbo Shang +20Benchmark DesignMulti-Turn Dialogue Evaluation

    Yuanbao Team, Tencent

  8. Holdout Best-of-N: Unbiased Evaluation and Its Cost

    Oct 6, 2026Shrey Shah, Yinheng LiBest-of-N SamplingPolicy Evaluation

  9. CriticHack: Evaluating Visual Rewards Under Robot Policy Optimization

    Oct 1, 2026Jiaxuan Luo, Xingguo Xu, Shanshan Wang +2RL for RoboticsReward Hacking

    Johns Hopkins University · Dalian University of Technology · Rutgers University +2

  10. When Residualization Helps an Audit: Format Effects, Slice Gains, and Their Limits

    Sep 21, 2026Daein Weon, Dong Ho KangLanguage Model Bias EvaluationLanguage Model Auditing

    Kookmin University · UStechlab

  11. Beyond Reference-Based Evaluation: Reward Models for Meta-Evaluation of Grammatical Error Correction

    Sep 18, 2026Ruotian Wu, Bill E. Johnson, Gene Saunders +3Grammatical Error CorrectionAutomated Evaluation

    University of Waterloo · Vector Institute · Scribendi Inc.

  12. F2^{2}DR: A Fine-Grained Full-Pipeline Reward Framework for DeepSearch Workflows

    Sep 17, 2026Bojian Xiong, Wentao Ding, Yujing Lu +11Reward ModelingLLM Evaluation

    Tianjin University · TJUNLP Lab, Tianjin University, Tianjin, China · Baidu Inc. +1

  13. TryOnReward: Learning Foveated Consistency for Reinforcement Fine-Tuning of Virtual Try-On

    Sep 6, 2026Xueheng Li, Yong Liu, Xiaolong Fu +5Virtual Try-OnMultimodal Reward Modeling

    JD.com, China · University of Science and Technology of China, China · South China University of Technology, China

  14. UniRRM: Unified Reasoning Reward Models Across Languages and Evaluation Paradigms

    Sep 5, 2026Peng Lai, Yichao Du, Junchao Wu +6Reward ModelingReward Model Evaluation

    Work done during an internship at Alibaba Cloud. · Southern University of Science and Technology · Wuhan University +4

  15. PRM-as-a-Judge 1.5: A Toolkit for Robot Process Assessment

    Aug 14, 2026Yuyang Liu, Yanqing Shen, Ruike Chen +22Robotic Policy EvaluationRobot Manipulation Benchmarks

  16. Evaluator Ensembles Under Reward Hacking: Covariance Geometry and Finite-Search Guarantees

    Aug 8, 2026Fariya Afrin, Ibne Farabi ShihabReward HackingLLM Evaluation

    Department of Computer Science, Kalinga Institute of Industrial Technology · Department of Computer Science, Iowa State University

  17. OSReward: Instituting Standardized Evaluation for Cross-Platform Computer-Use Reward Models

    Jul 30, 2026Qiushi Sun, Kanzhi Cheng, Yian Wang +20VLM EvaluationComputer-Use Agent Benchmarks

    The University of Hong Kong · Nanjing University · National University of Singapore +4

  18. SeekJudge: A Practical Reward Framework for Reinforcement Learning in Computer-Use Agents

    Jul 25, 2026Yang Wan, Zhenhao Zhang, Jierui Wang +1Reward ModelingLLM-as-a-Judge

    1Zhejiang University · University of Rochester

  19. Discretizing Reward Models

    Jun 19, 2026Vijay Viswanathan, Shiqi Wang, Devamanyu Hazarika +4Reward ModelingReinforcement Learning

    1Carnegie Mellon University · 2Meta Superintelligence Labs

  20. Beyond Rubrics: Exploration-Guided Evaluation Skills for Reward Modeling

    Jun 5, 2026Xing Yue, Linjuan Wu, Daoxin Zhang +2Reward ModelingLLM Evaluation

    Zhejiang University · Xiaohongshu Inc.

  21. Skill-RM: Unifying Heterogeneous Evaluation Criteria via Agent Skill

    Jun 2, 2026Tao Chen, Gangwei Jiang, Pengyu Cheng +10Reward ModelingReward Model Evaluation

    1Qwen Large Model Application Team, Alibaba · 2Sun Yat-sen University · 3The Chinese University of Hong Kong +3

  22. When RLHF Fails: A Mechanistic Taxonomy of Reward Hacking, Collapse, and Evaluator Gaming

    Jun 2, 2026Zelalem Abahana, David Evans, Satish Mahadevan Srinivasan +1Reward HackingLLM Evaluation

    College of Engineering, The Pennsylvania State University, University Park, PA, USA · Department of Computer Science, University of Virginia, Charlottesville, VA, USA · 3Penn State Great Valley, The Pennsylvania State University, Malvern, PA, USA +1

  23. HARVE: Hacking-Aware Reward-Head Vector Editing for Robust Reward Models

    Jun 2, 2026Shuang Liu, Yuxuan Bo, Qiuyang Zhao +4Reward ModelingReward Hacking

    Carnegie Mellon University · University of Virginia · Harvard University +4

  24. Expected Value Alignment for Generative Reward Modeling in Formal Mathematics Verification

    May 31, 2026Shihao Ji, Haotao Tan, Zihui Song +1Reward ModelingAutomated Theorem Proving

  25. Position: Good Embodied Reward Models Need Bad Behavior Data

    May 31, 2026Ran Tian, Yilin Wu, Andrea BajcsyReward ModelingPreference Alignment

    UC Berkeley · NVIDIA Research · Carnegie Mellon University

  26. EST-PRM: Stress-Testing Process Reward Models Before They Become Load-Bearing

    May 30, 2026Ibne Farabi Shihab, Fariya Afrin, Sanjeda Akter +1Process Reward ModelsLanguage Model Calibration

    Department of Computer Science, Iowa State University · Department of Computer Science, Kalinga Institute of Industrial Technology · Department of Civil, Construction & Environmental Engineering, Iowa State University

  27. FormalRewardBench: A Benchmark for Formal Theorem Proving Reward Models

    May 11, 2026Zeynel A. Uluşan, Burak S. Akbudak, Can S. Erer +1Reward ModelingAutomated Theorem Proving

    Koç University, Department of Computer Science and Engineering · Codeway Studios · Bo˘gaziçi University, Department of Computer Engineering +1

  28. Better Reward Models Are Not Always Less Biased: Social Alignment Beyond Standard Reward Benchmarks

    May 6, 2026Gayane Ghazaryan, Esra DönmezReward ModelingLanguage Model Safety Evaluation

    Institute for Natural Language Processing, University of Stuttgart · Interchange Forum for Reflecting on Intelligent Systems, University of Stuttgart

  29. StoryAlign: Evaluating and Training Reward Models for Story Generation

    May 6, 2026Haotian Xia, Hao Peng, Yunjia Qi +4Reward ModelingLanguage Model Generation Evaluation

    Department of Computer Science and Technology, Tsinghua University

  30. RMGAP: Benchmarking the Generalization of Reward Models across Diverse Preferences

    May 3, 2026Yangyang Zhou, Yi-Chen LiReward ModelingLLM Evaluation

    Beijing University of Posts and Telecommunications · National Key Laboratory for Novel Software Technology, Nanjing University, China · School of Artificial Intelligence, Nanjing University, China