Metraipolis › archive › dc_cf149eb97e75

Vision-Language Models Integrate Conflicting Evidence Like Biased Bayesian Observers — arXiv:2610.19407

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.19407
License (as archived): CC0-1.0
Archived: 2026-10-15T20:02:36.811Z
SHA-256 of the archived text: 44d896c520ef0c0d59e634dde3711558e4a40513494e830a285631a3b580beae
Stable URL of this page: https://metraipolis.com/documento/dc_cf149eb97e75

How to cite

Vision-Language Models Integrate Conflicting Evidence Like Biased Bayesian Observers — arXiv:2610.19407 — arXiv (en), https://arxiv.org/abs/2610.19407 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_cf149eb97e75 (Metraipolis archive dc_cf149eb97e75, sha256 44d896c520ef0c0d59e634dde3711558e4a40513494e830a285631a3b580beae)

BibTeX

@misc{dccf149eb97e75,
  title = {Vision-Language Models Integrate Conflicting Evidence Like Biased Bayesian Observers — arXiv:2610.19407},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.19407},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_cf149eb97e75; sha256 44d896c520ef0c0d59e634dde3711558e4a40513494e830a285631a3b580beae},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.19407. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Vision-language models must reconcile conflicting text and images. Prior work has characterized the resulting behavior in terms of modality biases--e.g., a "blind faith in text"--but why preferences vary across models, domains, and input conditions remains unclear. We develop a predictive account from a Bayesian perspective, positing that VLMs combine conditionally independent evidence from each modality with prior beliefs, subject to modality-specific biases. The resulting two-parameter cue combination model accurately predicts VLM preferences under unseen combinations of input degradation and transfers to new forms of unreliability without refitting. Separating evidence strength from integration bias explains how VLMs can follow text even while weighting visual evidence more heavily, reconciling reports of text and vision dominance under a single combination rule.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2610.19407 — 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 2026-10-15T19:35:48.337Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).