Metraipolis › archive › dc_cb4737579f21
MIRA: Reinforcing Multimodal Reasoning via Deceptive Contextual Augmentation — arXiv:2611.03087
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.03087
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-03T10:58:50.238Z
- SHA-256 of the archived text
75f412ce94d768a239f6df5a34f223a00512894e5e54ea322a8cca606db59f6a- Stable URL of this page
https://metraipolis.com/documento/dc_cb4737579f21
How to cite
MIRA: Reinforcing Multimodal Reasoning via Deceptive Contextual Augmentation — arXiv:2611.03087 — arXiv (en), https://arxiv.org/abs/2611.03087 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_cb4737579f21 (Metraipolis archive dc_cb4737579f21, sha256 75f412ce94d768a239f6df5a34f223a00512894e5e54ea322a8cca606db59f6a)
BibTeX
@misc{dccb4737579f21,
title = {MIRA: Reinforcing Multimodal Reasoning via Deceptive Contextual Augmentation — arXiv:2611.03087},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.03087},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_cb4737579f21; sha256 75f412ce94d768a239f6df5a34f223a00512894e5e54ea322a8cca606db59f6a},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.03087. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
While Reinforcement Learning with Verifiable Rewards (RLVR) has significantly advanced multimodal reasoning, existing frameworks suffer from a critical "perception-reasoning gap." Due to an over-reliance on seemingly relevant visual tokens, even minor perceptual perturbations can propagate through the reasoning process, leading to compounding hallucinations. To address this issue, we propose MIRA, a training framework that improves robustness to erroneous visual contexts while enhancing logical reasoning ability. By injecting filtered, deceptive visual contexts during training, we construct challenging "hard traps" that stress-test the model's robustness to misleading information. Furthermore, we introduce a group-wise reflection-triggering mechanism that enables the model to autonomously detect and rectify inconsistencies between visual cues and logical constraints, which is ultimately internalized as an intrinsic policy behavior. Importantly, MIRA does not rely on human annotations. Experimental results across multiple reasoning and perception benchmarks show that MIRA yields an average 7% improvement on in-domain benchmarks, with consistent improvements on vision-centric benchmarks. These findings highlight that active error recovery is essential for robust multimodal reasoning. Comprehensive ablation studies and analyses further provide insights into why MIRA is effective. 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). Pagina del lavoro: https://arxiv.org/abs/2611.03087 — 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 2026-11-03T10:49:38.705Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).