arXiv:2608.16417cs.CL2026-08

动态调整检索与推理规模,提升长文档理解准确率

D2-ScaleAgent: Dual-Dimensional Scaling for Long Document Understanding

论文配图:D2-ScaleAgent: Dual-Dimensional Scaling for Long Document Understanding
图 1 · 摘自论文原文
  • 根据问题难度动态扩展检索或细化推理,实现双维度自适应
  • 在MMLongBench-Doc等数据集上超越现有方法,证据覆盖更完整
  • 适合需要精准理解长篇图文文档的复杂问答任务

多模态检索增强生成(RAG)是视觉丰富长文档理解的关键技术。现有方法逐步向多智能体系统演进:先基于查询检索相关页面,再迭代理解页面内容。但这些方法通常依赖固定流程,缺乏测试时动态调节计算的能力,常导致证据不足。为此,我们提出D2-ScaleAgent,一种引入双维度缩放范式的智能体框架。其核心是基于查询内在难度、由验证器驱动的动态路由循环,围绕持续更新的证据库(作为智能体的动态工作记忆)运行:当需扩展检索时,智能体向外路由(检索缩放),将查询分解为属性并行检索页面,再自适应裁剪以保证证据全面覆盖;当需细粒度推理时,智能体向内路由(推理缩放),动态选择不同粒度和数量的子智能体从页面中提取证据。最终,D2-ScaleAgent实现证据链的逻辑闭环。大量实验表明,该方法在MMLongBench-Doc、LongDocURL等长文档基准上表现优异。

原文摘要 · Abstract (English)

Multi-modal retrieval-augmented generation (RAG) is a key technique for visually rich long document understanding. Existing multi-modal RAG methods are progressively advancing toward multi-agent systems: they first retrieve relevant pages based on a query, and then iteratively understand information within those pages. However, these methods typically rely on fixed workflows and lack the ability to dynamically scale computation at test time, often leading to insufficient evidence. To address this, we propose D2-ScaleAgent, an agentic framework that introduces a dual-dimensional scaling paradigm for retrieval and reasoning. The core of D2-ScaleAgent is a Verifier agent-driven dynamic routing loop based on the intrinsic difficulty of the query, centered around a continuously updated evidence bank that serves as the agent's dynamic working memory: when retrieval needs to be expanded, the agent routes outward (retrieval scaling), decomposing the query into attributes and performing parallel page retrieval, followed by adaptive pruning to ensure comprehensive evidence coverage. When fine-grained reasoning is required, the agent routes inward (reasoning scaling), dynamically selecting sub-agents with varying granularity and count to extract evidence from pages. Finally, D2-ScaleAgent achieves logical closure over the evidence chain. Extensive experiments demonstrate that D2-ScaleAgent is effective on long and visually rich document benchmarks like MMLongBench-Doc, LongDocURL, etc.

长文档理解多智能体动态缩放

Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品;中文卡片由大模型生成,请以原文为准。