Metraipolis › archive › dc_f6fd8ff3ffc1

RePair: Turning Retrieval Failures into Counterfactual Hard Pairs — arXiv:2608.29604

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.29604
License (as archived)
CC0-1.0
Archived
2026-10-17T09:44:16.935Z
SHA-256 of the archived text
c9a181f2f8e2d1c498f3bfbe66834696f0e8f385a1ebb994a1c7e8986e8200da
Stable URL of this page
https://metraipolis.com/documento/dc_f6fd8ff3ffc1

How to cite

RePair: Turning Retrieval Failures into Counterfactual Hard Pairs — arXiv:2608.29604 — arXiv (en), https://arxiv.org/abs/2608.29604 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_f6fd8ff3ffc1 (Metraipolis archive dc_f6fd8ff3ffc1, sha256 c9a181f2f8e2d1c498f3bfbe66834696f0e8f385a1ebb994a1c7e8986e8200da)
BibTeX
@misc{dcf6fd8ff3ffc1,
  title = {RePair: Turning Retrieval Failures into Counterfactual Hard Pairs — arXiv:2608.29604},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.29604},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_f6fd8ff3ffc1; sha256 c9a181f2f8e2d1c498f3bfbe66834696f0e8f385a1ebb994a1c7e8986e8200da},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.29604. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Vision-language retrieval with CLIP-style dual encoders achieves strong cross-modal performance, yet practical accuracy often hinges on localized semantic distinctions where top-ranked near misses differ from the true match by a single critical detail. Hard-sample mining can select confusable candidates but cannot construct corrected counterparts; synthetic augmentation can generate novel samples but, without conditioning on actual model failures, targets irrelevant dimensions of hardness. We observe that a top-ranked false positive is a counterfactual scaffold---sharing most of the query's semantics while differing in a localized failure-causing residual. Minimally correcting this residual yields a hard positive of the ground truth in the same modality; the corrected and unedited versions form a hard negative pair that straddles the decision boundary, producing complementary pull--push supervision. We introduce RePair, guided by three principles---Validity, Minimality, and Locality---which mines false positives bidirectionally, applies LLM-guided counterfactual editing, and trains with a local hard-pair contrastive objective. On Flickr30K and COCO30K, RePair outperforms controlled augmentation baselines with only 107K synthetic samples---26\%--75\% fewer than comparable methods---confirming failure-conditioned repair is more data-efficient than error-agnostic augmentation.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-01 in cs.IR).
Categorie arXiv: Information Retrieval (cs.IR) ; Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2608.29604 — 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 2026-10-17T09:12:53.470Z dalla pagina https://arxiv.org/catchup/cs.IR/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).