Metraipolis › archive › dc_47950deb2676

DocPilot: Interactive Agent Orchestration and Training for Long Visual Document Understanding — arXiv:2610.21327

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.21327
License (as archived)
CC0-1.0
Archived
2026-10-16T09:03:01.292Z
SHA-256 of the archived text
9fe09e9934ff4491635b15c79205da09f46a90d82f7815e03b1cf22ef3ddee00
Stable URL of this page
https://metraipolis.com/documento/dc_47950deb2676

How to cite

DocPilot: Interactive Agent Orchestration and Training for Long Visual Document Understanding — arXiv:2610.21327 — arXiv (en), https://arxiv.org/abs/2610.21327 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_47950deb2676 (Metraipolis archive dc_47950deb2676, sha256 9fe09e9934ff4491635b15c79205da09f46a90d82f7815e03b1cf22ef3ddee00)
BibTeX
@misc{dc47950deb2676,
  title = {DocPilot: Interactive Agent Orchestration and Training for Long Visual Document Understanding — arXiv:2610.21327},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.21327},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_47950deb2676; sha256 9fe09e9934ff4491635b15c79205da09f46a90d82f7815e03b1cf22ef3ddee00},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.21327. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Multimodal long-document question answering requires models to seek sparse evidence across pages and modalities, understand heterogeneous document content, and reason over the collected evidence. Even advanced vision-language models (VLMs) can overlook sparse evidence and struggle with cross-modal reasoning in a single pass. Existing document agents further introduce multi-step interaction, but some still rely on predefined workflows or coarse-grained actions, limiting the flexibility and granularity of document reasoning. We introduce DocPilot, an agentic document framework that supports expressive orchestration and makes complex document solving learnable by small language models (SLMs). DocPilot orchestrates evidence seeking through page-level navigation and fine-grained evidence localization, and integrates local visual inspection, code-based numerical reasoning, and cross-modal reasoning for multimodal understanding. To transfer these capabilities to small models, we propose curriculum-based agentic training that progresses from single-ability learning to adaptive orchestration through supervision from data-grounded reasoning traces and executable interaction trajectories. Across MMLongBench-Doc, DocBench, FinRAGBench-V, and MADQA, DocPilot achieves strong gains with frontier VLM backbones, while DocPilot-4B delivers competitive performance against substantially larger models, demonstrating strong effectiveness and generalization across diverse long-document reasoning settings. The code is available at this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2610.21327 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-16T08:44:17.728Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).