Metraipolis › archive › dc_97de620b5159
Revisiting Certified Defense with Differential Privacy on Vision Transformers — arXiv:2609.31310
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.31310
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-08T02:09:24.384Z
- SHA-256 of the archived text
0e142debe9b899481de862458a884844d5036b7bffb4502b5e9b7be8797ee595- Stable URL of this page
https://metraipolis.com/documento/dc_97de620b5159
How to cite
Revisiting Certified Defense with Differential Privacy on Vision Transformers — arXiv:2609.31310 — arXiv (en), https://arxiv.org/abs/2609.31310 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_97de620b5159 (Metraipolis archive dc_97de620b5159, sha256 0e142debe9b899481de862458a884844d5036b7bffb4502b5e9b7be8797ee595)
BibTeX
@misc{dc97de620b5159,
title = {Revisiting Certified Defense with Differential Privacy on Vision Transformers — arXiv:2609.31310},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.31310},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_97de620b5159; sha256 0e142debe9b899481de862458a884844d5036b7bffb4502b5e9b7be8797ee595},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.31310. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Certified defenses that incorporate differential privacy have proven effective on Convolutional Neural Networks (CNNs), furnishing rigorous robustness guarantees against norm-bounded adversaries. However, the certified robustness behavior of Pixel Differential Privacy (PixelDP) remains largely unexplored with the self-attention architecture now dominating the deep-learning landscape. Given that the Transformer has a profound impact on our daily applications from the digital world to the physical world, it is crucial to study certified robustness through differential-privacy-style stability. To fill this research gap, we revisit this construction in Vision Transformers and identify a failure mode that is largely hidden in the convolutional setting. When noise is injected after the patch embedding, the Laplace mechanism with the inherited grouped $\ell_1$ sensitivity bound collapses to chance-level accuracy across noise scales, whereas the Gaussian mechanism remains trainable. This contrast isolates the source of failure: not the injected noise itself, but the geometry of the sensitivity constraint. We show that the attenuation induced by the inherited $\Delta_{1,1}$ projection increases with layer width and kernel size according to a random-matrix scale $C/(\sqrt{M}+\sqrt{N})$. Replacing the $\ell_1$-type constraint with a spectral-norm constraint eliminates the collapse across datasets and architectures, but creates a fundamental obstacle: the repaired models no longer satisfy the sensitivity condition required by the standard Laplace certificate. We resolve this mismatch by deriving a dimension-free $(\varepsilon,\ \delta)$-privacy guarantee for the Laplace mechanism under $\ell_2$ sensitivity through concentration of the privacy loss.
Tema degli esploratori: fiducia digitale e crittografia (retro-catalogo arXiv: annunci del 2026-09-28 in cs.CR).
Categorie arXiv: Cryptography and Security (cs.CR) ; Applications (stat.AP).
Pagina del lavoro: https://arxiv.org/abs/2609.31310 — 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 2026-10-08T01:42:25.553Z dalla pagina https://arxiv.org/catchup/cs.CR/2026-09-28?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).