一句话：提出两阶段工具交互训练框架 ToolTrain，用拒绝采样监督微调与强化学习提升 LLM 使用检索工具进行问题定位的能力。
问题定位旨在找出解决软件问题所需修改的代码位置，但自然语言问题描述与缺陷代码间的语义鸿沟要求沿代码依赖进行多跳推理。现有基于 LLM 的智能体虽引入仓库检索工具，仍难以在多步推理中有效使用这些工具。本文提出 ToolTrain，一个两阶段工具交互训练框架，结合拒绝采样监督微调与工具交互强化学习，增强 LLM 利用检索工具进行问题定位的能力。实验显示，训练后模型在同规模 LLM 中达到最优，32B 模型在函数级定位上甚至超过 Claude-3.7-Sonnet；定位性能提升也转化为端到端问题解决性能提升，表明面向问题定位的训练是改善自动化软件开发的有效策略。
Issue localization, the process of identifying code locations that need modification to resolve software issues, is a critical yet challenging task in software development. The semantic gap between natural language issue descriptions and faulty code requires complex multi-hop reasoning through code dependencies. Existing LLM-based agents attempt to address this by integrating repository retrieval tools, which poses a higher demand for LLMs to effectively utilize various tools during multi-step reasoning for issue localization. To tackle this challenge, we present ToolTrain, a two-stage tool-interactive training framework combining rejection-sampled supervised fine-tuning and tool-interactive reinforcement learning to enhance LLMs’ ability to use retrieval tools for issue localization. Experimental results show that ToolTrain-trained models achieve state-of-the-art performance among same-size LLMs, with our 32B model even surpassing Claude-3.7-Sonnet on function-level localization. The results also show that improved localization performance translates to better end-to-end issue resolution performance. This further demonstrates that training for issue localization is a viable and effective strategy for improving automated software development.
@inproceedings{paperbot11793,
title = {Enhancing Issue Localization Agent with Tool-Interactive Training},
author = {Zexiong Ma and Chao Peng and Qunhong Zeng and Pengfei Gao and Yanzhen Zou and Bing Xie},
booktitle = {ICSE},
year = {2026},
doi = {10.1145/3744916.3773137}
}