acceptodds
Desk rejected at ICLR 2027

Towards More Reliable Adversarial Robustness Evaluation via Sequential Difference Maximization

Desk rejected at ICLR 2027, so left out of the paper lists and the map.

Abstract

Gradient-based adversarial attacks are important methods for evaluating model robustness. However, since the proposal of APGD, it has been difficult for such methods to achieve significant breakthroughs. To achieve such an effect, we first analyze the issue of "high-loss non-adversarial examples" that degrades attack performance in previous methods, and prove that this issue arises from inappropriate objectives for adversarial example generation. Subsequently, we reconstruct the objective as "maximizing the difference between the non-ground-truth label probability upper bound and the ground-truth label probability", and propose a novel and powerful gradient-based attack method named Sequential Difference Maximization (SDM). SDM establishes a three-layer optimization framework of "Cycle-Stage-Step". It adopts the negative probability loss function and the Directional Probability Difference Ratio (DPDR) loss function in the initial and subsequent optimization stages, respectively, and approaches the ideal objective of adversarial example generation via stage-wise sequential optimization. Experiments demonstrate that compared with previous state-of-the-art methods, SDM not only achieves stronger attack performance but also exhibits superior cost-effectiveness.

Primary area:
alignment, fairness, safety, privacy, and societal considerations
Keywords:
deep learning security, adversarial example, adversarial attack, robustness evaluation, image classification

Related papers

Discussion (0)

Sign in to comment.