all InfoSec news
On the Duality Between Sharpness-Aware Minimization and Adversarial Training
Feb. 26, 2024, 5:11 a.m. | Yihao Zhang, Hangzhou He, Jingyu Zhu, Huanran Chen, Yifei Wang, Zeming Wei
cs.CR updates on arXiv.org arxiv.org
Abstract: Adversarial Training (AT), which adversarially perturb the input samples during training, has been acknowledged as one of the most effective defenses against adversarial attacks, yet suffers from a fundamental tradeoff that inevitably decreases clean accuracy. Instead of perturbing the samples, Sharpness-Aware Minimization (SAM) perturbs the model weights during training to find a more flat loss landscape and improve generalization. However, as SAM is designed for better clean accuracy, its effectiveness in enhancing adversarial robustness remains …
adversarial arxiv aware cs.ai cs.cr cs.lg duality math.oc minimization training
More from arxiv.org / cs.CR updates on arXiv.org
Jobs in InfoSec / Cybersecurity
Security Engineer
@ Celonis | Munich, Germany
Security Engineer, Cloud Threat Intelligence
@ Google | Reston, VA, USA; Kirkland, WA, USA
IT Security Analyst*
@ EDAG Group | Fulda, Hessen, DE, 36037
Scrum Master/ Agile Project Manager for Information Security (Temporary)
@ Guidehouse | Lagunilla de Heredia
Waste Incident Responder (Tanker Driver)
@ Severn Trent | Derby , England, GB
Risk Vulnerability Analyst w/Clearance - Colorado
@ Rothe | Colorado Springs, CO, United States