{"ID":2857108,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.10197","arxiv_id":"2510.10197","title":"Don't Just Fine-tune the Agent, Tune the Environment","abstract":"Large Language Model (LLM) agents show great promise for complex, multi-turn tool-use tasks, but their development is often hampered by the extreme scarcity of high-quality training data. Supervised fine-tuning (SFT) on synthetic data leads to overfitting, whereas standard reinforcement learning (RL) struggles with a critical cold-start problem and training instability. To address these challenges, we introduce $\\textbf{Environment Tuning}$, a novel training paradigm that enables agents to learn complex behaviors directly from problem instances without relying on pre-collected expert trajectories. $\\textbf{Environment Tuning}$ orchestrates this learning process through a structured curriculum, actionable environment augmentation that provides corrective feedback, and fine-grained progress rewards to ensure stable and efficient exploration. Using only 400 problem instances from Berkeley Function-Calling Leaderboard (BFCL) benchmark, our method not only achieves competitive in-distribution performance against strong baselines but also demonstrates superior out-of-distribution generalization, overcoming the performance collapse common to SFT-based approaches. Our work presents a paradigm shift from supervised fine-tuning on static trajectories to dynamic, environment-based exploration, paving the way for training more robust and data-efficient agents. The code is available at https://github.com/inclusionAI/AWorld-RL/tree/main/EnvTuning.","short_abstract":"Large Language Model (LLM) agents show great promise for complex, multi-turn tool-use tasks, but their development is often hampered by the extreme scarcity of high-quality training data. Supervised fine-tuning (SFT) on synthetic data leads to overfitting, whereas standard reinforcement learning (RL) struggles with a c...","url_abs":"https://arxiv.org/abs/2510.10197","url_pdf":"https://arxiv.org/pdf/2510.10197v2","authors":"[\"Siyuan Lu\",\"Zechuan Wang\",\"Hongxuan Zhang\",\"Qintong Wu\",\"Leilei Gan\",\"Chenyi Zhuang\",\"Jinjie Gu\",\"Tao Lin\"]","published":"2025-10-11T12:35:15Z","proceeding":"cs.AI","tasks":"[\"cs.AI\"]","methods":"[\"Reinforcement Learning\",\"Large Language Model\",\"Language Model\",\"LoRA\"]","has_code":false,"code_links":[{"ID":608415,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2857108,"paper_url":"https://arxiv.org/abs/2510.10197","paper_title":"Don't Just Fine-tune the Agent, Tune the Environment","repo_url":"https://github.com/inclusionAI/AWorld-RL","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}