{"ID":2847943,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.26270","arxiv_id":"2510.26270","title":"Graph-Enhanced Policy Optimization in LLM Agent Training","abstract":"Multi-step LLM agents in interactive environments represent a crucial step toward long-horizon decision-making. To train such agents, group-based reinforcement learning is widely adopted, which reinforces trajectories with higher relative performance within the group. However, in most existing methods, every step within a trajectory and every trajectory with the same terminal reward receive identical credit, regardless of their actual contributions. Since different states play different structural roles in an online state-transition graph built from sampled trajectories, their impacts should be differentiated and converted into task-aware credit at both the step and trajectory levels. We therefore present Graph-Enhanced Policy Optimization (GEPO), a framework for dual-level structural credit assignment in multi-step LLM agent training. Specifically, GEPO derives a state-level Task-Conditioned Criticality score that combines topological betweenness on the state-transition graph with semantic similarity to the task prompt. Based on this score, trajectory-level credit is reshaped through a state-adaptive discount, while step-level credit is scaled by the criticality of its successor state. Experimental results show that GEPO outperforms the strongest baselines by 1.1\\% in success rate on ALFWorld, 3.2\\% on WebShop, and 3.8\\% on average across search-augmented QA tasks at the 7B scale. Compared with flat group-based methods, GEPO reduces across-seed variance and concentrates gradient signals on the most critical steps.","short_abstract":"Multi-step LLM agents in interactive environments represent a crucial step toward long-horizon decision-making. To train such agents, group-based reinforcement learning is widely adopted, which reinforces trajectories with higher relative performance within the group. However, in most existing methods, every step withi...","url_abs":"https://arxiv.org/abs/2510.26270","url_pdf":"https://arxiv.org/pdf/2510.26270v2","authors":"[\"Jiazhen Yuan\",\"Zhike Gong\",\"Jinquan Hang\",\"Zhengbiao Bai\",\"Wei Zhao\"]","published":"2025-10-30T08:53:41Z","proceeding":"cs.AI","tasks":"[\"cs.AI\"]","methods":"[\"Reinforcement Learning\",\"Large Language Model\"]","has_code":false}