Metraipolis › archive › dc_33ab107d5cd8

TailProp: content-adaptive light- and heavy-tailed propagation for vision — arXiv:2609.11081

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.11081
License (as archived)
CC0-1.0
Archived
2026-10-16T05:55:15.806Z
SHA-256 of the archived text
0504bde718217e5f110c7761d90ae0899127388ce390698d53b4a67a016bd40a
Stable URL of this page
https://metraipolis.com/documento/dc_33ab107d5cd8

How to cite

TailProp: content-adaptive light- and heavy-tailed propagation for vision — arXiv:2609.11081 — arXiv (en), https://arxiv.org/abs/2609.11081 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_33ab107d5cd8 (Metraipolis archive dc_33ab107d5cd8, sha256 0504bde718217e5f110c7761d90ae0899127388ce390698d53b4a67a016bd40a)
BibTeX
@misc{dc33ab107d5cd8,
  title = {TailProp: content-adaptive light- and heavy-tailed propagation for vision — arXiv:2609.11081},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.11081},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_33ab107d5cd8; sha256 0504bde718217e5f110c7761d90ae0899127388ce390698d53b4a67a016bd40a},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.11081. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Science-inspired vision models show that explicit propagation dynamics can provide structured and interpretable alternatives to conventional token mixing. Existing formulations, however, typically construct and adapt visual propagation within a particular dynamical family, while visual representations can require substantially different spatial interactions across samples, channels, and network stages. We explore cross-regime adaptive propagation and introduce TailProp, a hierarchical vision backbone built upon the Tail Propagation Operator (TPO). TPO uses Gaussian and Cauchy stable-process propagators as complementary bases with rapidly decaying and heavy-tailed spatial influence, and predicts a content-conditioned channel-wise coefficient to adaptively combine them. Because this coefficient is spatially shared, the two responses are fused directly in the DCT domain with a single DCT/IDCT pair, yielding $O(N^{1.5})$ spatial mixing for square feature maps with $N=HW$ and fixed channel width. Across image classification, object detection, semantic segmentation, robustness, and cross-backbone restoration, TailProp consistently outperforms matched propagation baselines; TailProp-B reaches 84.4% Top-1 accuracy on ImageNet-1K, 50.3/44.8 box/mask AP under the 3x Mask R-CNN schedule, and 50.8% mIoU on ADE20K. Controlled ablations further show that these gains are not explained by single-basis propagation, an additional same-family branch, or within-family adaptive order alone, supporting complementary two-basis propagation as an effective design principle for visual representation learning.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-11 in cs.LG).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.11081 — 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 2026-10-16T05:44:25.066Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-11?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).