| title | InSight-doc: Agentic Visual Perception for Long-Document Understanding |
|---|---|
| authors | Kaican Li, Weiyan Xie, Lewei Yao, Jiannan Wu, Lanqing Hong, Yongxiang Huang, Nevin L. Zhang |
| date | 2026-08-11 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2608.10628 |
| upvotes | 6 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf code |
| abstract | Long-document understanding often requires reasoning over many visually rich pages, making inference costly and prone to context rot. In this work, we propose InSight-doc, an agentic visual perception framework that treats visual resolution as an adaptive reasoning-time resource. InSight-doc starts from low resolution and selectively zooms into high-resolution regions for finer evidence, without relying on any external retriever. To train such an agent, we construct an active-perception corpus of 17.9K high-quality SFT examples with region-level zoom-in trajectories, accompanied by 19.2K hard RL examples. Through SFT+RL, InSight-doc-8B improves the baseline by 4.3--16.4 accuracy points over document VQA benchmarks. On long documents, it reduces hallucination by more than 40% and inference latency by 41%--68% while maintaining an accuracy lead. Our code, datasets, and model are released at https://github.com/m-Just/InSight-doc . |
cite
@misc{li2026insight,
title={InSight-doc: Agentic Visual Perception for Long-Document Understanding},
author={Kaican Li and Weiyan Xie and Lewei Yao and Jiannan Wu and Lanqing Hong and Yongxiang Huang and Nevin L. Zhang},
year={2026},
eprint={2608.10628},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2608.10628}
}