Metraipolis › archive › dc_a620bfb0d3ee

Evaluating Cross-Signal Coherence in Mammography Vision-Language Models — arXiv:2611.12574

Source: arXiv (en)
Original: https://arxiv.org/abs/2611.12574
License (as archived): CC0-1.0
Archived: 2026-11-10T18:45:12.315Z
SHA-256 of the archived text: 16ccaf27313c5f4009eb2e8429673b859937b722c1fa1fc9fe5a2308b8f6c668
Stable URL of this page: https://metraipolis.com/documento/dc_a620bfb0d3ee

How to cite

Evaluating Cross-Signal Coherence in Mammography Vision-Language Models — arXiv:2611.12574 — arXiv (en), https://arxiv.org/abs/2611.12574 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a620bfb0d3ee (Metraipolis archive dc_a620bfb0d3ee, sha256 16ccaf27313c5f4009eb2e8429673b859937b722c1fa1fc9fe5a2308b8f6c668)

BibTeX

@misc{dca620bfb0d3ee,
  title = {Evaluating Cross-Signal Coherence in Mammography Vision-Language Models — arXiv:2611.12574},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.12574},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a620bfb0d3ee; sha256 16ccaf27313c5f4009eb2e8429673b859937b722c1fa1fc9fe5a2308b8f6c668},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.12574. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

As vision-language models (VLMs) move toward interpreting medical images directly, their outputs increasingly include not just a label but a stated confidence and a written justification. Yet the field still predominantly evaluates VLMs using just accuracy. We evaluate the coherence between answer, confidence, and explanation across four VLMs on breast density and BI-RADS categorization, using two mammography datasets. We perform controlled perturbations of the question, multiple choice option orders, and the images and assess changes in the three outputs. Our results suggest confidence does not decrease even as the mammography image is changed. Shuffling the options also changes most VLMs' answers. Notably, there are cases where a model's own stated reasoning contradicts its final answer, a distinction accuracy alone cannot capture but one that governs whether an AI-generated answer can be trusted.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Computation and Language (cs.CL); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2611.12574 — 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 2026-11-10T18:37:52.486Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).