Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2601.22055 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914291924336640 |
|---|---|
| author | Du, Yaxin Song, Junru Zhou, Yifan Wang, Cheng Gu, Jiahao Chen, Zimeng Chen, Menglan Yao, Wen Yang, Yang Wen, Ying Chen, Siheng |
| author_facet | Du, Yaxin Song, Junru Zhou, Yifan Wang, Cheng Gu, Jiahao Chen, Zimeng Chen, Menglan Yao, Wen Yang, Yang Wen, Ying Chen, Siheng |
| contents | Retrieval-augmented generation is a practical paradigm for question answering over long documents, but it remains brittle for multimodal reading where text, tables, and figures are interleaved across many pages. First, flat chunking breaks document-native structure and cross-modal alignment, yielding semantic fragments that are hard to interpret in isolation. Second, even iterative retrieval can fail in long contexts by looping on partial evidence or drifting into irrelevant sections as noise accumulates, since each step is guided only by the current snippet without a persistent global search state. We introduce $G^2$-Reader, a dual-graph system, to address both issues. It evolves a Content Graph to preserve document-native structure and cross-modal semantics, and maintains a Planning Graph, an agentic directed acyclic graph of sub-questions, to track intermediate findings and guide stepwise navigation for evidence completion. On VisDoMBench across five multimodal domains, $G^2$-Reader with Qwen3-VL-32B-Instruct reaches 66.21\% average accuracy, outperforming strong baselines and a standalone GPT-5 (53.08\%). |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_22055 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | $G^2$-Reader: Dual Evolving Graphs for Multimodal Document QA Du, Yaxin Song, Junru Zhou, Yifan Wang, Cheng Gu, Jiahao Chen, Zimeng Chen, Menglan Yao, Wen Yang, Yang Wen, Ying Chen, Siheng Computation and Language Retrieval-augmented generation is a practical paradigm for question answering over long documents, but it remains brittle for multimodal reading where text, tables, and figures are interleaved across many pages. First, flat chunking breaks document-native structure and cross-modal alignment, yielding semantic fragments that are hard to interpret in isolation. Second, even iterative retrieval can fail in long contexts by looping on partial evidence or drifting into irrelevant sections as noise accumulates, since each step is guided only by the current snippet without a persistent global search state. We introduce $G^2$-Reader, a dual-graph system, to address both issues. It evolves a Content Graph to preserve document-native structure and cross-modal semantics, and maintains a Planning Graph, an agentic directed acyclic graph of sub-questions, to track intermediate findings and guide stepwise navigation for evidence completion. On VisDoMBench across five multimodal domains, $G^2$-Reader with Qwen3-VL-32B-Instruct reaches 66.21\% average accuracy, outperforming strong baselines and a standalone GPT-5 (53.08\%). |
| title | $G^2$-Reader: Dual Evolving Graphs for Multimodal Document QA |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2601.22055 |