The paper rigorously analyzes the statistical and learning-theoretic properties of adversarial training models for classification, focusing on empirical optimal partial transport. It establishes two central limit theorems—one centered at the expected empirical value and another at the population value with smoothing—by leveraging the uniqueness of optimal potentials across various optimal transport formulations and empirical process theory. In the binary setting, the authors prove uniqueness of the optimal potential via a connection to multi-marginal optimal transport, and as additional results they derive stability of the saddle point, sample complexity, and concentration bounds for generalization error.
By Kim Jakwang, Kwon Dohyun
arXiv:2506.12454v2 Announce Type: replace-cross
Abstract: What fundamentally distinguishes an adversarial attack from a misclassification due to limited model expressivity or finite data? In this wor...
By Matteo Vilucchio, Lenka Zdeborov\'a, Bruno Loureiro
arXiv:2410. 07719v4 Announce Type: replace Abstract: Despite being widely adopted as a canonical framework for learning robust models, adversarial training suffers from robust overfitting.
By Yuelin Xu, Xiao Zhang
arXiv:2606. 27685v1 Announce Type: cross Abstract: Pervasive data contamination -- stemming from measurement errors, outliers, or adversarial corruption -- has motivated the development of robust statistical methods.
By Shixiang Liu, Hanming Yang
arXiv:2606. 00320v1 Announce Type: new Abstract: We present an online, distribution-free framework for controlling the Conditional Value-at-Risk (CVaR), extending conformal tail risk control to non-stationary and adversarial environments.
By Catherine Chen, Jingyan Shen, Zhun Deng, Lihua Lei
arXiv:2608. 09688v1 Announce Type: cross Abstract: Adversarial training under long tailed distributions suffers from a dual imbalance: the class imbalance skews the training objective toward head classes, and the adversarial inner maximization may further amplify this bias.
By Mengnan Zhao, Geyong Min, Lihe Zhang, Tianhang Zheng, Jie Cui
Adversarial training under long tailed distributions suffers from a dual imbalance: the class imbalance skews the training objective toward head classes, and the adversarial inner maximization may further amplify this bias. Existing methods mitigate this issue by correcting class priors or adapting class wise robust supervision, yet they treat each class in isolation and fail to identify which boundaries drive long tailed collapse.
The paper investigates how deterministic Gradient Descent behaves when applied to Adversarial Training with large, constant step sizes. It analyzes both a linear logistic regression setting—examining where parameters converge (implicit bias) and how fast losses decrease (optimization)—and neural network training, focusing on the evolution of loss sharpness. The study reveals three key distinctions from Standard Training: the implicit bias gap persists even with large steps, there is no uniform convergence rate for optimization error, and sharpness follows a two‑stage dynamic unlike the Edge‑of‑Stability pattern seen in Standard Training.
By Yi Feng, Andrea Paudice
arXiv:2610.00861v1 Announce Type: cross
Abstract: Adversarial optimization under a shared $\ell_1$ budget requires deciding not only how much perturbation to use, but also where that limited budget s...
By Melika Shirian, Kianoosh Vadaei
The paper introduces Inverse Knowledge Distillation (IKD), an attack‑agnostic technique that enhances adversarial transferability by maximizing the discrepancy between benign and adversarial prediction distributions on a surrogate model. IKD employs a CE/KL‑equivalent soft‑label objective to push adversarial predictions away from a fixed benign anchor, leveraging Fisher‑sensitive surrogate directions. The authors provide theoretical analysis showing CE and KL induce identical gradients, derive a lower bound on Fisher‑subspace overlap, and demonstrate through extensive ImageNet experiments that IKD consistently improves black‑box attack performance across CNN, ViT, and defended models.
By Wenyuan Wu, Yuan Sun, Yingke Chen, Chao Su, Xi Peng, Dezhong Peng, Xu Wang
The paper introduces two methods to improve adversarial training by targeting the model’s confidence structure. One method blends cross‑entropy with KL divergence to simultaneously degrade accuracy and distribution consistency, while the other uses Fisher geometry to generate perturbations that maximize distribution distance on the confidence manifold. Experiments show these techniques outperform existing approaches in robustness and transferability, especially for high‑confidence structural perturbations.
By Chenlong Wei, Hanling Wang, Chong Zhang, Qinkai Yu, Fangyu Wu, Xiaobo Jin
arXiv:2608. 11815v1 Announce Type: new Abstract: Transfer-based adversarial attacks craft adversarial examples using surrogate models to mislead black-box victim models.
By Yaohua Liu, Yifan Guo, Jiaxin Gao