{"ID":2829298,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2512.12658","arxiv_id":"2512.12658","title":"CogDoc: Towards Unified thinking in Documents","abstract":"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.","short_abstract":"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 l...","url_abs":"https://arxiv.org/abs/2512.12658","url_pdf":"https://arxiv.org/pdf/2512.12658v1","authors":"[\"Qixin Xu\",\"Haozhe Wang\",\"Che Liu\",\"Fangzhen Lin\",\"Wenhu Chen\"]","published":"2025-12-14T12:14:17Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[\"Reinforcement Learning\"]","has_code":false}