mlsanityData to 7 Oct 2026

CogDoc: Towards Unified thinking in Documents

arXiv 2512.12658 ↗First posted 14 Dec 2025cs.CVcited by 3

Not on the map. Why?

Abstract (arXiv)

Current document reasoning paradigms are constrained by a fundamental trade-off between scalability (processing long-context documents) and fidelity (capturing fine-grained, multimodal details). To bridge this gap, we propose CogDoc, a unified coarse-to-fine thinking framework that mimics human cognitive processes: a low-resolution "Fast Reading" phase for scalable information localization,followed by a high-resolution "Focused Thinking" phase for deep reasoning. We conduct a rigorous investigation into post-training strategies for the unified thinking framework, demonstrating that a Direct Reinforcement Learning (RL) approach outperforms RL with Supervised Fine-Tuning (SFT) initialization. Specifically, we find that direct RL avoids the "policy conflict" observed in SFT. Empirically, our 7B model achieves state-of-the-art performance within its parameter class, notably surpassing significantly larger proprietary models (e.g., GPT-4o) on challenging, visually rich document benchmarks.

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 37 of its 47 references that are in the map. First posted on arXiv 14 Dec 2025.

Builds on

37 of 47 references.

References · 37

  1. 2020Dense Passage Retrieval for Open-Domain Question Answeringcited by 3,779
  2. 2023Retrieval-Augmented Generation for Large Language Models: A Surveycited by 2,185
  3. 2023MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGIcited by 2,044
  4. 2024HybridFlow: A Flexible and Efficient RLHF Frameworkcited by 1,980
  5. 2023Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflectioncited by 1,800
  6. 2025InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Modelscited by 1,769
  7. 2024LlamaFactory: Unified Efficient Fine-Tuning of 100+ Language Modelscited by 1,442
  8. 2023REPLUG: Retrieval-Augmented Black-Box Language Modelscited by 786
  9. 2019LayoutLM: Pre-training of Text and Layout for Document Image Understandingcited by 549
  10. 2024RAPTOR: Recursive Abstractive Processing for Tree-Organized Retrievalcited by 519
  11. 2022LayoutLMv3: Pre-training for Document AI with Unified Text and Image Maskingcited by 481
  12. 2025VL-Rethinker: Incentivizing Self-Reflection of Vision-Language Models with Reinforcement Learningcited by 315
25 more
  1. 2025Pixel Reasoner: Incentivizing Pixel-Space Reasoning with Curiosity-Driven Reinforcement Learningcited by 267
  2. 2024VisRAG: Vision-based Retrieval-augmented Generation on Multi-modality Documentscited by 237
  3. 2025Multimodal Chain-of-Thought Reasoning: A Comprehensive Surveycited by 176
  4. 2023SlideVQA: A Dataset for Document Visual Question Answering on Multiple Imagescited by 175
  5. 2022Improving Passage Retrieval with Zero-Shot Question Generationcited by 164
  6. 2024mPLUG-DocOwl2: High-resolution Compressing for OCR-free Multi-page Document Understandingcited by 113
  7. 2023DocLLM: A layout-aware generative language model for multimodal document understandingcited by 112
  8. 2024M3DocRAG: Multi-modal Retrieval is What You Need for Multi-page Multi-document Understandingcited by 110
  9. 2024Unifying Multimodal Retrieval via Document Screenshot Embeddingcited by 107
  10. 2022Hierarchical multimodal transformers for Multi-Page DocVQAcited by 102
  11. 2024MMed-RAG: Versatile Multimodal RAG System for Medical Vision Language Modelscited by 84
  12. 2025ReFocus: Visual Editing as a Chain of Thought for Structured Image Understandingcited by 81
  13. 2025Agent0: Unleashing Self-Evolving Agents from Zero Data via Tool-Integrated Reasoningcited by 68
  14. 2025ViDoRAG: Visual Document Retrieval-Augmented Generation via Dynamic Iterative Reasoning Agentscited by 64
  15. 2025Emergent Hierarchical Reasoning in LLMs through Reinforcement Learningcited by 42
  16. 2025MMedAgent-RL: Optimizing Multi-Agent Collaboration for Multimodal Medical Reasoningcited by 37
  17. 2025Reverse-Engineered Reasoning for Open-Ended Generationcited by 32
  18. 2024SV-RAG: LoRA-Contextualizing Adaptation of MLLMs for Long Document Understandingcited by 25
  19. 2025To Code or not to Code? Adaptive Tool Integration for Math Language Models via Expectation-Maximizationcited by 25
  20. 2025DualRAG: A Dual-Process Approach to Integrate Reasoning and Retrieval for Multi-Hop Question Answeringcited by 20
  21. 2025Docopilot: Improving Multimodal Models for Document-Level Understandingcited by 19
  22. 2020Learning Context-aware Task Reasoning for Efficient Meta-reinforcement Learningcited by 11
  23. 2025PeRL: Permutation-Enhanced Reinforcement Learning for Interleaved Vision-Language Reasoningcited by 11
  24. 2025PixelCraft: A Multi-Agent System for High-Fidelity Visual Reasoning on Structured Imagescited by 11
  25. 2025HKRAG: Holistic Knowledge Retrieval-Augmented Generation Over Visually-Rich Documentscited by 5

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 1; 2026: 2.

3 papers.

  1. 2026Bad Seeing or Bad Thinking? Rewarding Perception for Multimodal Reasoningcited by 6
  2. 2026Doc-V*:Coarse-to-Fine Interactive Visual Reasoning for Multi-Page Document VQAcited by 4
  3. 2025From Chaos to Clarity: Schema-Constrained AI for Auditable Biomedical Evidence Extraction from Full-Text PDFscited by 1