Metraipolis › archive › dc_9969a83eca26

Visual Input and Its Framing Affect Attribute-based Descriptions Produced by Large Vision-Language Models — arXiv:2609.18345

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.18345
License (as archived)
CC0-1.0
Archived
2026-10-15T16:16:27.167Z
SHA-256 of the archived text
7c9d763ff7e1cf34d430ab1c9b1fb8aac53e7bc8f6f2e8fdc49b62d21fa5a1c5
Stable URL of this page
https://metraipolis.com/documento/dc_9969a83eca26

How to cite

Visual Input and Its Framing Affect Attribute-based Descriptions Produced by Large Vision-Language Models — arXiv:2609.18345 — arXiv (en), https://arxiv.org/abs/2609.18345 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_9969a83eca26 (Metraipolis archive dc_9969a83eca26, sha256 7c9d763ff7e1cf34d430ab1c9b1fb8aac53e7bc8f6f2e8fdc49b62d21fa5a1c5)
BibTeX
@misc{dc9969a83eca26,
  title = {Visual Input and Its Framing Affect Attribute-based Descriptions Produced by Large Vision-Language Models — arXiv:2609.18345},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.18345},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_9969a83eca26; sha256 7c9d763ff7e1cf34d430ab1c9b1fb8aac53e7bc8f6f2e8fdc49b62d21fa5a1c5},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.18345. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Large vision-language models (LVLMs) are commonly used with only a single text prompt as the input, or plus an image. In this paper, we demonstrate that when the image exists, even if the text prompt is not about the specific instance (but only the concept it belongs to) in that image, the response would still be affected. For example, when the text prompt only asks for the attribute descriptions of a dog breed, an image depicting a specific dog from that breed would shift the response. Further, how the specific instance is framed in that image would determine towards which the response shifts. Detailed analyses also reveal that in the response, physical terms increase from 18% for text-only to 45% (40%) for subject-focused (subject-in-situation) framings. Overall, the unexpected effects of visual cues on LVLMs highlight the need to understand the presence of an image and its framing when evaluating the robustness of LVLMs.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-17 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.18345 — 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 2026-10-15T15:45:19.778Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-17?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).