Metraipolis › archive › dc_afffa8666606
Seeing Isn't Enough: Unlocking Visual Reasoning by Bridging the Reasoning-Access Gap with TIGER — arXiv:2611.10552
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.10552
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-10T18:41:03.166Z
- SHA-256 of the archived text
9a0c9333d6b2f6b79142bf25959821524ede2486d2e01b5974c190ce3291b25b- Stable URL of this page
https://metraipolis.com/documento/dc_afffa8666606
How to cite
Seeing Isn't Enough: Unlocking Visual Reasoning by Bridging the Reasoning-Access Gap with TIGER — arXiv:2611.10552 — arXiv (en), https://arxiv.org/abs/2611.10552 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_afffa8666606 (Metraipolis archive dc_afffa8666606, sha256 9a0c9333d6b2f6b79142bf25959821524ede2486d2e01b5974c190ce3291b25b)
BibTeX
@misc{dcafffa8666606,
title = {Seeing Isn't Enough: Unlocking Visual Reasoning by Bridging the Reasoning-Access Gap with TIGER — arXiv:2611.10552},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.10552},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_afffa8666606; sha256 9a0c9333d6b2f6b79142bf25959821524ede2486d2e01b5974c190ce3291b25b},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.10552. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Multimodal Large Language Models (MLLMs) inherit strong textual reasoning capabilities from their language backbones and text pretraining, but often fail to apply the same level of competency to visual inputs. We show that this degradation cannot be attributed to only perception failures, by evaluating models on semantically-equivalent text and rendered-text image inputs: models can accurately read text in the images and recover near-text performance when provided their own transcriptions, but just cannot reason well on the images directly. We characterize this discrepancy as a modality reasoning-access gap, where models fail to route perceived visual content into existing reasoning machinery accessible with text. To bridge this, we propose TIGER (Text-to-Image Gap-targeted training for Enhanced Reasoning), which first mines synthetic modality counterfactuals where a model succeeds on a text problem but fails when provided the corresponding rendered image, and post-trains the model on the image input via Group Relative Policy Optimization (GRPO), along with SFT and DPO variants. Importantly, although trained exclusively on synthetic rendered-text images with no multimodal annotations, TIGER results in model performance improvements across diverse real-world visual benchmarks including geometric diagrams (MathVerse), multi-disciplinary sciences (EMMA, MMMU), complex charts (ChartQAPro), and embodied spatial reasoning (RoboSpatial-Home). TIGER also complements existing state-of-the-art RLVR visual reasoning methods, yielding further performance gains. Analysis involving reasoning-subspace activation and activation patching provide support indicating that TIGER helps to align visual-to-reasoning pathways toward text reasoning states. Our implementation 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) ; Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2611.10552 — 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 2026-11-10T18:37:37.494Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).