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

2026-08-17

AI Summary

This study addresses the limitation of fixed workflows in existing multimodal RAG systems, which often yield insufficient evidence for long-document understanding. We propose a Dual-Dimensional Expansion Agent framework that introduces a novel paradigm for dynamically scaling both retrieval and reasoning. By constructing a validator-driven adaptive routing mechanism and dynamic evidence memory, the system expands retrieval breadth or refines reasoning depth on demand to establish a closed logical evidence chain. Evaluations on visually rich long-document benchmarks, including MMLongBench-Doc, demonstrate that this framework significantly improves comprehension accuracy for complex queries. Consequently, our approach effectively resolves the critical challenge of aligning evidence acquisition with appropriate reasoning granularity in multimodal long-context tasks.

Abstract

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.

Problem

• Long Document Understanding
• Multi-modal RAG
• Test-time Scaling
• Evidence Sufficiency

Innovation

• Dual-Dimensional Scaling
• Dynamic Routing
• Verifier Agent
• Multi-modal RAG
• Evidence Bank