发表机构
Wuhan University; Microsoft（武汉大学; 微软）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究提出DocAtlas系统，将长文档理解视为可变状态交互，通过可变文档工具结合自改进检索等技术，在MMLongBench-Doc上用GPT-5.4达71.4%，还可提升紧凑型VLM智能体性能。
AI 中文摘要
长文档理解需要模型跨多个页面、布局、表格、图形和图表查找并整合证据。现有的检索增强系统通常在生成前从静态索引中选择证据，而近期的智能体系统虽增加了多轮工具使用，但往往依赖行为由提示设定的冻结专有主干模型。我们提出DocAtlas，一种将长文档理解视为可变状态信息寻求过程的系统。我们将DocAtlas实例化为可变文档工具：一个外部环境，用于决定每一步搜索、读取、存储、审查和向模型展示的文档信息。给定文档和问题，该工具提供搜索、读取、记笔记和审查工具，维护分层树和笔记存储，并在智能体记录证据时更新两者。DocAtlas在固定上下文预算下结合自改进检索、选择性证据访问和主动工作记忆。同一工具支持大型VLMs的推理时使用，以及针对紧凑型VLM智能体的端到端强化学习。使用GPT-5.4，DocAtlas在MMLongBench-Doc上达到71.4%，超过人类专家参考的65.8%。在DocAtlas环境中经端到端RL训练的Qwen3.5-4B VLM达到63.7%，相比直接输入基线的54.4%，表明可变文档工具设计可大幅提升紧凑型文档智能体的性能。
英文摘要
Long-document understanding requires models to find and combine evidence across many pages, layouts, tables, figures, and charts. Existing retrieval-augmented systems usually select evidence from a static index before generation, while recent agentic systems add multi-turn tool use but often rely on frozen proprietary backbones whose behavior is set by prompts. We present DocAtlas, a system that treats long-document understanding as a mutable-state information-seeking process. We instantiate DocAtlas as a mutable document harness: an external environment that determines what document information is searched, read, stored, reviewed, and shown to the model at each step. Given a document and question, the harness exposes search, reading, note-taking, and review tools, maintains a hierarchical tree and note store, and updates both as the agent records evidence. DocAtlas combines self-improving retrieval, selective evidence access, and active working memory under a fixed context budget. The same harness supports inference-time use with large VLMs and end-to-end reinforcement learning for compact VLM agents. With GPT-5.4, DocAtlas reaches 71.4\% on MMLongBench-Doc, exceeding the human-expert reference of 65.8\%. A Qwen3.5-4B VLM trained with end-to-end RL in the DocAtlas environment reaches 63.7\%, compared with a 54.4\% direct-input baseline, showing that mutable document-harness design can improve compact document agents by a large margin.
