Metraipolis › archive › dc_dedff5281d4a
GRIT: Gradient Masking for Visually Misgrounded Tokens in Text-Rich Reasoning — arXiv:2610.41629
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.41629
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-03T05:33:32.212Z
- SHA-256 of the archived text
8dc0ead7851dc45e65e330c917eebde18ceb72aef544eaa853cef5218eb76966- Stable URL of this page
https://metraipolis.com/documento/dc_dedff5281d4a
How to cite
GRIT: Gradient Masking for Visually Misgrounded Tokens in Text-Rich Reasoning — arXiv:2610.41629 — arXiv (en), https://arxiv.org/abs/2610.41629 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_dedff5281d4a (Metraipolis archive dc_dedff5281d4a, sha256 8dc0ead7851dc45e65e330c917eebde18ceb72aef544eaa853cef5218eb76966)
BibTeX
@misc{dcdedff5281d4a,
title = {GRIT: Gradient Masking for Visually Misgrounded Tokens in Text-Rich Reasoning — arXiv:2610.41629},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.41629},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_dedff5281d4a; sha256 8dc0ead7851dc45e65e330c917eebde18ceb72aef544eaa853cef5218eb76966},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.41629. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Reinforcement learning with verifiable rewards (RLVR) has recently been extended to vision-language models (VLMs), yet its effectiveness remains limited for text-rich reasoning tasks. We highlight a key factor contributing to this limitation: models frequently arrive at correct final answers despite misperceiving visual information during intermediate steps. We define this phenomenon as Correct-but-Visually-Misgrounded Reasoning (CVMR). Because standard RLVR rewards only final-answer correctness, it fails to differentiate between properly grounded and misgrounded reasoning, reinforcing both equally. Through token-level analysis, we reveal that CVMR tokens exhibit distinct attention patterns that differ systematically from well-grounded ones. Based on this observation, we propose GRIT, a training framework that incorporates visual grounding into RLVR to mitigate the impact of CVMR. Furthermore, to address the lack of text-rich reasoning benchmarks that allow for systematic evaluation across text length and reasoning depth, we introduce GRIT-Bench. Experiments demonstrate that GRIT significantly improves both reasoning accuracy and grounding fidelity over existing RLVR methods. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI); Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2610.41629 — 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 2026-11-03T05:18:52.222Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).