Mitigating catastrophic overfitting in fast adversarial training via dynamic polyak weight averaging and gradient norm dual-penalty
Neurocomputing · 2026-05 · 678 · 133149
DOI: 10.1016/j.neucom.2026.133149 ↗Kurz gefasst
Fachartikel in der Zeitschrift Neurocomputing, 2026-05. Verfasst von Anjum Iqbal und 4 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 48 Literaturangaben.
Abstract (Originalsprache)
Kein Abstract verfügbar.
Zitieren
APA
Iqbal, A., Kong, W., Iqbal, Y., Zhang, M., & Hussain, I. (2026). Mitigating catastrophic overfitting in fast adversarial training via dynamic polyak weight averaging and gradient norm dual-penalty. Neurocomputing, 678, 133149. https://doi.org/10.1016/j.neucom.2026.133149
Vancouver
Iqbal A, Kong W, Iqbal Y, Zhang M, Hussain I. Mitigating catastrophic overfitting in fast adversarial training via dynamic polyak weight averaging and gradient norm dual-penalty. Neurocomputing. 2026;678:133149. doi:10.1016/j.neucom.2026.133149
BibTeX
@article{Iqbal2026,
author = {Iqbal, Anjum and Kong, Weiqiang and Iqbal, Yasir and Zhang, Maozhen and Hussain, Ijaz},
title = {Mitigating catastrophic overfitting in fast adversarial training via dynamic polyak weight averaging and gradient norm dual-penalty},
journal = {Neurocomputing},
year = {2026},
volume = {678},
pages = {133149},
doi = {10.1016/j.neucom.2026.133149}
}
EUVIMED