@inproceedings{zheng-etal-2026-doc,
    title = "Doc-$V^*$: Coarse-to-Fine Interactive Visual Reasoning for Multi-Page Document {VQA}",
    author = "Zheng, Yuanlei  and
      Fu, Pei  and
      Li, Hang  and
      Wang, Ziyang  and
      Zhang, Yuyi  and
      Ruan, Wenyu  and
      Zhang, Xiaojin  and
      Wei, Zhongyu  and
      Luo, Zhenbo  and
      Luan, Jian  and
      Chen, Wei  and
      Bai, Xiang",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.acl-long.2129/",
    pages = "45901--45923",
    ISBN = "979-8-89176-390-6",
    abstract = "Multi-page Document Visual Question Answering requires reasoning over semantics, layouts, and visual elements in long, visually dense documents. Existing OCR-free methods face a trade-off between capacity and precision: end-to-end models scale poorly with document length, while visual retrieval-based pipelines are brittle and passive. We propose \textbf{Doc-$V^*$}, an \textbf{OCR-free agentic} framework that casts multi-page DocVQA as sequential evidence aggregation. \textbf{Doc-$V^*$} begins with a thumbnail overview, then actively navigates via semantic retrieval and targeted page fetching, and aggregates evidence in a structured working memory for grounded reasoning. Trained by imitation learning from expert trajectories and further optimized with Group Relative Policy Optimization, \textbf{Doc-$V^*$} balances answer accuracy with evidence-seeking efficiency. Across five benchmarks, \textbf{Doc-$V^*$} outperforms open-source baselines and approaches proprietary models, improving out-of-domain performance by up to \textbf{47.9{\%}} over RAG baseline. Other results reveal effective evidence aggregation with selective attention, not increased input pages."
}