Locating Failure in Multi-Page Visually Rich Document Understanding Systems: An Empirical Attribution
cs.AI
Submitted: 2026-08-08
Updated: 2026-10-15
License: http://creativecommons.org/licenses/by/4.0/
The gist: Multi-page visually-rich document understanding (MP-VRDU) requires managing evidence that is sparse, spread across pages, and often exceeds a model's context window.
Terminology
Abstract
Multi-page visually-rich document understanding (MP-VRDU) requires managing evidence that is sparse, spread across pages, and often exceeds a model's context window. Prior work has produced competing, largely untested claims about how these systems should be built. We attribute incorrect answers to three failure modes, representation, selection, and reasoning, and isolate each over a multi-page document understanding dataset by intervening on one while holding the others fixed. We find that vision is necessary but does not replace text extraction, that missing pages bound accuracy while distractors cost little, and that reasoners fail to integrate evidence across pages even when it is fully supplied. Prompting can shift reasoning behaviour substantially, improving some outcomes at the expense of others. We translate these findings into guidance for building such systems under a fixed compute budget.
Sources
- Qwen3-VL Technical Report
- M3DocRAG: Multi-modal Retrieval is What You Need for Multi-page Multi-document Understanding
- PaddleOCR-VL: Boosting Multilingual Document Parsing via a 0.9B Ultra-Compact Vision-Language Model
- A Survey on MLLM-based Visually Rich Document Understanding: Methods, Challenges, and Emerging Trends
- MDocAgent: A Multi-Modal Multi-Agent Framework for Document Understanding
- DocSLM: A Small Vision-Language Model for Long Multimodal Document Understanding
- A Survey of Large Language Models Attribution
- Resolving Evidence Sparsity: Agentic Context Engineering for Long-Document Understanding
- MinerU2.5: A Decoupled Vision-Language Model for Efficient High-Resolution Document Parsing
- UNIDOC-BENCH: A Unified Benchmark for Document-Centric Multimodal RAG
- Unlimited OCR Works
- EfficientLLM: Efficiency in Large Language Models
- The Reasoning Bottleneck in Graph-RAG: Structured Prompting and Context Compression for Multi-Hop QA
- Doc-V*:Coarse-to-Fine Interactive Visual Reasoning for Multi-Page Document VQA
- DocLens : A Tool-Augmented Multi-Agent Framework for Long Visual Document Understanding
- InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems