Reinforcement learning (RL) for large language model (LLM) agents remains challenging in long-horizon tasks with sparse outcome rewards, where assigning credit to individual actions is difficult. Methods such as GRPO assign the same trajectory-level advantage to all actions, while critic-based methods require learning an additional value function that can be costly and unstable. We introduce LLM Value Function RL (LVRL), which uses the pretrained knowledge and in-context learning ability of an LLM to estimate policy-aware state values without finetuning a separate critic. LVRL conditions the value model on group context and calibrates its predictions using known values at the initial and terminal states, enabling step-level credit assignment from estimated value differences. We further extend this idea to learning from an expert policy through on-policy distillation (OPD), introducing LLM Value Function OPD (LVOPD), which focuses teacher supervision on actions associated with decreases in estimated value. We evaluate our approach across diverse agentic benchmarks: AppWorld, BabyAI, and SWE-bench Verified. Across policy scales (4B to 32B), LVRL consistently and substantially outperforms both GRPO and PPO. On AppWorld and SWE-bench Verified, LVRL yields absolute improvements of up to 13.6 and 8.6 percentage points over GRPO. Notably, on tasks requiring extended exploration like BabyAI, LVRL bridges the gap from total failure to 75.0% success. Furthermore, LVOPD yields consistent improvements over standard distillation baselines on AppWorld. Finally, Monte-Carlo analysis confirms that our frozen critic accurately identifies semantic pitfalls and recoveries, proving that LVRL extracts robust, policy-aware step-level credit from sparse terminal outcomes.
Towards Practical Credit Assignment for Deep Reinforcement Learning
Yandex（俄罗斯搜索引擎公司（或：雅虎迪））；Samsung AI Center NYC（三星人工智能中心纽约分部）；Artificial Intelligence Research Institute (AIRI)（人工智能研究所（AIRI））；National Research University Higher School of Economics（高等经济学院）
RACA: Relation-Aware Credit Assignment for Ad-Hoc Cooperation in Multi-Agent Deep Reinforcement Learning
University of Chinese Academy of Sciences（中国科学院大学）；Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）；CAS Center for Excellence in Brain Science and Intelligence Technology（中国科学院脑科学与智能技术卓越创新中心）
Sequence Compression Speeds Up Credit Assignment in Reinforcement Learning
The Swiss AI Lab IDSIA, USI & SUPSI（瑞士IDSIA人工智能实验室、瑞士意大利语区大学、瑞士南部应用科学大学）；University of Alberta（阿尔伯塔大学）；Alberta Machine Intelligence Institute（阿尔伯塔机器智能研究所）；AI Initiative, King Abdullah University of Science and Technology（阿卜杜拉国王科技大学人工智能计划）
ASCT: Attentive Search over Counterfactual Trees for Credit Assignment in Agentic Reinforcement Learning
University of Hong Kong（香港大学）；The Chinese University of Hong Kong（香港中文大学）；Jiangxi Science and Technology Normal University（江西科技师范大学）；Shenzhen University（深圳大学）
Stop Summation: Min-Form Credit Assignment Is All Process Reward Model Needs for Reasoning
State Key Laboratory of Multimodal Artificial Intelligence Systems, Institute of Automation, Chinese Academy of Sciences（多模态人工智能系统国家重点实验室，自动化研究所，中国科学院）；School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）；Shanghai Artificial Intelligence Laboratory（上海人工智能实验室）；Tencent（腾讯）；Faculty of Innovation Engineering, Macau University of Science and Technology（澳门科技大学创新工程学院）；DeSci Center of Parallel Intelligence, Obuda University（并行智能德Sci中心，奥布达大学）；Research Center for Chinese Economics and Social Security, University of Chinese Academy of Sciences（中国经济社会安全研究中心，中国科学院大学）；Institute of Automation, Chinese Academy of Sciences（自动化研究所，中国科学院）
Would I have gotten that reward? Long-term credit assignment by counterfactual contribution analysis
Google DeepMind（谷歌DeepMind）；Department of Computer Science, ETH Zürich（苏黎世联邦理工学院计算机科学系）；Princeton Neuroscience Institute, Princeton University（普林斯顿大学普林斯顿神经科学研究所）；Department of Psychology, Princeton University（普林斯顿大学心理学系）
