Metraipolis › archive › dc_3779133e387f

Neither Adversarial Training Nor Purification: Emergent Adversarial Robustness from Oscillatory Predictive Learning — arXiv:2609.08683

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.08683
License (as archived)
CC0-1.0
Archived
2026-10-16T13:58:22.409Z
SHA-256 of the archived text
de5819a8737ef87808c4a5cc121be0ccdcdd596d3981e53290874c274539162c
Stable URL of this page
https://metraipolis.com/documento/dc_3779133e387f

How to cite

Neither Adversarial Training Nor Purification: Emergent Adversarial Robustness from Oscillatory Predictive Learning — arXiv:2609.08683 — arXiv (en), https://arxiv.org/abs/2609.08683 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3779133e387f (Metraipolis archive dc_3779133e387f, sha256 de5819a8737ef87808c4a5cc121be0ccdcdd596d3981e53290874c274539162c)
BibTeX
@misc{dc3779133e387f,
  title = {Neither Adversarial Training Nor Purification: Emergent Adversarial Robustness from Oscillatory Predictive Learning — arXiv:2609.08683},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.08683},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3779133e387f; sha256 de5819a8737ef87808c4a5cc121be0ccdcdd596d3981e53290874c274539162c},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.08683. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Adversarial robustness in computer vision is still largely achieved through adversarial training or test-time adversarial purification, both of which introduce significant computational overhead by generating adversarial examples during training or performing iterative denoising at test time. We study whether empirical robustness can instead emerge from architectural and representation-learning inductive biases. We introduce Oscillatory Predictive Learning (OPL), a two-stage framework that combines Artificial Kuramoto Oscillatory Neurons (AKOrN) with predictive self-supervised pretraining using X-PhiNet. Because our default checkpoint uses randomized initial oscillator states, we compare it with other randomized adversarial defense methods that provide precise, reproducible, and strong attack protocols. Experiments on CIFAR-10 and CIFAR-100, with additional corruption evaluation on CIFAR-10-C, demonstrate that our method achieves competitive results under the AutoAttack-rand evaluation protocol. On CIFAR-10 and CIFAR-100, OPL attains 76.63$\pm$0.76$\%$ and 50.44$\%$ robust accuracy, respectively, under $\ell_\infty$, $\epsilon=8/255$, AutoAttack-rand with EoT $K=20$.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-09 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.08683 — 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 2026-10-16T13:39:43.060Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-09?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).