MidTool：面向智能体工具使用的训练中期数据合成
MidTool: Mid-training Data Synthesis for Agentic Tool Use
浏览论文内容
中文总结 AI 辅助
本研究提出MidTool流水线，构建面向智能体工具使用的训练中期数据，在MidTool-Mix上微调Qwen3模型，可显著提升工具使用相关下游任务性能，证明通用工具使用需专门训练中期调整。
中文摘要 AI 辅助
训练中期正被日益视为塑造大语言模型能力的关键阶段。近期研究表明，针对性的训练中期调整可强化数学、科学等依赖推理的能力，还能提升软件工程场景下的智能体能力。本研究探讨了并行但较少被探索的智能体能力：通用工具使用。我们提出MidTool，这是一个用于智能体工具使用训练中期的开放语料库构建流水线，它将大规模网页、PDF及代码数据，与来自真实世界工具API、MCP技能、基于文档的工作流的合成监督相结合。MidTool旨在教会模型识别工具可供性、从上下文提炼论据、组合工具调用工作流，以及从不完整信息中恢复。我们在MidTool-Mix上对Qwen3-4B-Base和Qwen3-8B-Base进行训练中期调整，随后结合监督微调（SFT）和强化学习（RL）进行后续的训练后调整。与基线方法相比，MidTool-Mix在BFCL、tau2-Bench和MCP Universe上，无论在SFT还是RL设置下，均能持续提升下游性能。这些结果表明，通用工具使用与其他重要的大语言模型能力一样，受益于专门的训练中期调整，而非完全依赖训练后调整。
英文摘要
Mid-training is increasingly recognized as a critical stage for shaping the capabilities of large language models. Recent work has shown that targeted mid-training can strengthen reasoning-intensive abilities such as math and science, and can also improve agentic capabilities in software-engineering settings. In this work, we study the parallel but less explored agentic capability: general tool use. We present MidTool, an open corpus construction pipeline for agentic tool-use mid-training that combines large-scale web, PDF, and code data with synthesized supervision from real-world tool APIs, MCP skills, and document-grounded workflows. MidTool is designed to teach models how to recognize tool affordances, ground arguments from context, compose tool call workflow, and recover from incomplete information. We mid-train Qwen3-4B-Base and Qwen3-8B-Base on MidTool-Mix, and then apply follow-up post-training with both supervised fine-tuning and reinforcement learning. Compared with baselines, MidTool-Mix consistently improves downstream performance under both SFT and RL on BFCL, tau2-Bench, and MCP Universe. These results suggest that general tool use, like other important LLM capabilities, benefits from dedicated mid-training rather than being left entirely to post-training.
发表机构
- University of Washington（华盛顿大学）
- Snowflake（雪花公司）
- University of North Carolina at Chapel Hill（北卡罗来纳大学教堂山分校）
机构由 AI 辅助整理，请以论文原文为准。
