Metraipolis › archive › dc_11c322aee758

Assessing Representation Semantics Under Language Pretraining in Fine-Grained Classification — arXiv:2610.28051

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.28051
License (as archived)
CC0-1.0
Archived
2026-10-21T07:08:11.148Z
SHA-256 of the archived text
82d3e71855d410cddba582cedbf0a68823b04ea29734d417dd5028f29c657394
Stable URL of this page
https://metraipolis.com/documento/dc_11c322aee758

How to cite

Assessing Representation Semantics Under Language Pretraining in Fine-Grained Classification — arXiv:2610.28051 — arXiv (en), https://arxiv.org/abs/2610.28051 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_11c322aee758 (Metraipolis archive dc_11c322aee758, sha256 82d3e71855d410cddba582cedbf0a68823b04ea29734d417dd5028f29c657394)
BibTeX
@misc{dc11c322aee758,
  title = {Assessing Representation Semantics Under Language Pretraining in Fine-Grained Classification — arXiv:2610.28051},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.28051},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_11c322aee758; sha256 82d3e71855d410cddba582cedbf0a68823b04ea29734d417dd5028f29c657394},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.28051. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Understanding how supervision signals shape semantic visual representations remains challenging, particularly in fine-grained, low-data settings. Standard evaluations of vision models rely on downstream accuracy, which often fails to capture structural differences in learned representations. We propose a compact explainability-driven framework for comparing encoders trained with classification-only and caption-generation supervision on CUB-200. Using a controlled two-stage protocol in which both encoders are ultimately optimized for the same classification task, we enable a direct comparison of the resulting representations. Saliency maps are generated using attention rollout and gradient-weighted rollout and evaluated with five metrics covering faithfulness (positive and negative perturbation) and localization (keypoint emphasis, background attribution, and Earth Mover's Distance). Across random seeds and explanation methods, caption supervision yields representations that are more sensitive to the removal of salient regions, better aligned with annotated object parts, and less reliant on background regions, whereas classification supervision produces representations that are slightly more robust to the removal of low-saliency regions. An ablation study on the Oxford-IIIT Pet dataset shows consistent trends across key metrics. These results demonstrate that the proposed framework makes supervision-induced representational differences observable and quantifiable beyond downstream accuracy alone. More broadly, it provides a systematic basis for assessing learned representations and comparing alternative training and transfer-learning strategies.

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/2610.28051 — 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 2026-10-21T07:05:43.612Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).