Metraipolis › archive › dc_74b570f3d239
Curvature Under Attack in hZACH-ViT: Gauge Symmetry, Boundary Saturation, and Adversarial Failure — arXiv:2610.00680
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.00680
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-06T19:16:26.908Z
- SHA-256 of the archived text
1248aac8ca47581b95a2ad5b5b00cccf13ddf287691e7178ad97d87f0e0f29bd- Stable URL of this page
https://metraipolis.com/documento/dc_74b570f3d239
How to cite
Curvature Under Attack in hZACH-ViT: Gauge Symmetry, Boundary Saturation, and Adversarial Failure — arXiv:2610.00680 — arXiv (en), https://arxiv.org/abs/2610.00680 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_74b570f3d239 (Metraipolis archive dc_74b570f3d239, sha256 1248aac8ca47581b95a2ad5b5b00cccf13ddf287691e7178ad97d87f0e0f29bd)
BibTeX
@misc{dc74b570f3d239,
title = {Curvature Under Attack in hZACH-ViT: Gauge Symmetry, Boundary Saturation, and Adversarial Failure — arXiv:2610.00680},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.00680},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_74b570f3d239; sha256 1248aac8ca47581b95a2ad5b5b00cccf13ddf287691e7178ad97d87f0e0f29bd},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.00680. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Curvature is often treated as an intrinsic property of a representation, although its empirical effect also depends on coordinate scale, learned logit temperature, and numerical safeguards. We study this interaction in hZACH-ViT, a compact Vision Transformer with Euclidean, Poincare, and spherical prototype heads. The backbone architecture, seed-specific initialization, 50-per-class training subset, and optimization protocol are matched across three MedMNIST datasets and five seeds. At the fixed comparison curvature $c=1$, Poincare has the lowest class-macro PGD attack-success rate in all 12 dataset-budget cells and under a stronger CE+DLR multi-restart attack on all three datasets, but it also has the lowest clean MacroF1. An end-to-end curvature intervention changes the interpretation. Reducing Poincare curvature to $c=0.1$ improves clean MacroF1 in every one of the 15 paired seed-dataset comparisons and removes hard boundary clipping, yet on OrganAMNIST it increases strong attack success from $89.7\%$ to $99.3\%$ (paired difference $+9.57$ points; 95\% hierarchical bootstrap CI $[+5.52,+14.03]$). At $c=1$, $40$-$47\%$ of clean Poincare features are hard-clipped, the radial Jacobian of the inherited map is nearly zero, and dimensionless attack trajectories are unusually long and inefficient. The spherical head provides a control: its curvature change is an exact scale gauge to floating-point precision and produces much smaller attack differences. These results do not establish intrinsic hyperbolic robustness. They identify an implementation-sensitive regime in which curvature, scale, and proximity to the Poincare boundary jointly organize clean recognition and adversarial representation motion. Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-10-02 in cs.CV). Categorie arXiv: Machine Learning (cs.LG) ; Computer Vision and Pattern Recognition (cs.CV). Pagina del lavoro: https://arxiv.org/abs/2610.00680 — 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 2026-10-06T17:37:36.927Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-10-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).