Metraipolis › archive › dc_f9bbd4881293

Conditional Visual Evidence Utility: State-Dependent Rank Reversals in Frozen Vision-Language Encoders — arXiv:2608.28316

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.28316
License (as archived)
CC0-1.0
Archived
2026-10-17T11:33:16.949Z
SHA-256 of the archived text
9c0c7b4fe53d710901307ac69a498d049ed1b4a00576666bb531a7bc4849b251
Stable URL of this page
https://metraipolis.com/documento/dc_f9bbd4881293

How to cite

Conditional Visual Evidence Utility: State-Dependent Rank Reversals in Frozen Vision-Language Encoders — arXiv:2608.28316 — arXiv (en), https://arxiv.org/abs/2608.28316 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_f9bbd4881293 (Metraipolis archive dc_f9bbd4881293, sha256 9c0c7b4fe53d710901307ac69a498d049ed1b4a00576666bb531a7bc4849b251)
BibTeX
@misc{dcf9bbd4881293,
  title = {Conditional Visual Evidence Utility: State-Dependent Rank Reversals in Frozen Vision-Language Encoders — arXiv:2608.28316},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.28316},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_f9bbd4881293; sha256 9c0c7b4fe53d710901307ac69a498d049ed1b4a00576666bb531a7bc4849b251},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.28316. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Static importance scores compress visual evidence into a single ranking, but the value of remaining evidence can change after one cue has been observed. We study this possibility in controlled compositional visual search, where color, shape, and texture evidence can be independently exposed and their conditional marginal utility measured across acquisition states. In a held-out confirmation on 800 scenes, frozen OpenCLIP and SigLIP exhibit robust state-dependent rank reversals that concentrate in candidate-overlap regimes designed to induce ordering changes, persist across two evidence-accumulation constructions and ten equivalent query wordings, and collapse to near-chance-scale behavior under query-scene derangement. A subsequent role-balanced follow-up on 1,200 scenes rotates the abstract roles of initially strong, redundancy-inducing, and comparator attributes; the positive-minus-negative reversal contrast remains positive across all 24 role-permutation, backbone, and evidence-mode cells, although residual attribute-identity effects remain. We further distinguish measured replanning opportunity from prospective predictability. Matched-first-action utility analyses show substantial opportunity to rerank remaining evidence, but lightweight predictors using posterior-based or acquired-embedding state representations do not establish a robust incremental advantage of acquired-state information over legal static controls on the role-balanced benchmark. Together, these results show that conditional visual evidence utility is reliably state dependent in this controlled setting, while separating the existence of changing utility from the stronger claim that those changes are prospectively predictable by a learned selector.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-08-31 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2608.28316 — 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 2026-10-17T11:10:05.328Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-08-31?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).