Metraipolis › archive › dc_6933ceef8a7d

GramLoop: Training-Free Gram-Gated Replay for Robust Dense Prediction — arXiv:2608.29113

Source: arXiv (en)
Original: https://arxiv.org/abs/2608.29113
License (as archived): CC0-1.0
Archived: 2026-10-17T09:53:32.609Z
SHA-256 of the archived text: 5b9193400c9d503c26f780ea9c8249ab552db0aa804064ae6cc303feaa388946
Stable URL of this page: https://metraipolis.com/documento/dc_6933ceef8a7d

How to cite

GramLoop: Training-Free Gram-Gated Replay for Robust Dense Prediction — arXiv:2608.29113 — arXiv (en), https://arxiv.org/abs/2608.29113 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_6933ceef8a7d (Metraipolis archive dc_6933ceef8a7d, sha256 5b9193400c9d503c26f780ea9c8249ab552db0aa804064ae6cc303feaa388946)

BibTeX

@misc{dc6933ceef8a7d,
  title = {GramLoop: Training-Free Gram-Gated Replay for Robust Dense Prediction — arXiv:2608.29113},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.29113},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_6933ceef8a7d; sha256 5b9193400c9d503c26f780ea9c8249ab552db0aa804064ae6cc303feaa388946},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.29113. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We aim to improve frozen DINOv3 dense-prediction models under distribution shift by adding inference computation inside the visual backbone, without changing model weights, task adapters, or prediction heads. The challenge is that repeated transformer-block computation must refine dense features without disrupting the pairwise patch relations that DINOv3 uses to preserve spatial structure. We introduce GramLoop, a training-free framework that replays a short transformer window and controls each replay through final-layer cosine-Gram consistency. Each proposal is propagated through the frozen suffix, measured against the standard DINOv3 trajectory, and accepted through a patchwise gate at the replay-window endpoint. Across object detection and semantic segmentation under corruptions, perturbations, and natural shifts, GramLoop improves all five shifted benchmarks over the paired DINOv3 baseline. On COCO-O, it improves mAP by +0.252 and Effective Robustness by +0.250, while preserving clean ADE20K performance. Code will be released at this https URL .

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-01 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2608.29113 — 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 2026-10-17T09:28:47.982Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).