Exposing Vulnerability Monoculture Through Systemic Adversarial Robustness↗ (원문, 새 창)

본문 미리보기

arXiv:2610.27353v1 Announce Type: new Abstract: Beneath the apparent diversity of foundation models lies an increasingly homogeneous technical stack of common training pipelines, distillation methods, and data. These shared algorithmic choices can expose an otherwise distinct model ecosystem to concentrated risks leading to a vulnerability monoculture. Yet current practice in assessing adversarial robustness is restricted to individual robustness scores and attack transfer rates, leaving open t