English

ATRAS: Adversarially Trained Robust Architecture Search

Machine Learning 2021-06-15 v1

Abstract

In this paper, we explore the effect of architecture completeness on adversarial robustness. We train models with different architectures on CIFAR-10 and MNIST dataset. For each model, we vary different number of layers and different number of nodes in the layer. For every architecture candidate, we use Fast Gradient Sign Method (FGSM) to generate untargeted adversarial attacks and use adversarial training to defend against those attacks. For each architecture candidate, we report pre-attack, post-attack and post-defense accuracy for the model as well as the architecture parameters and the impact of completeness to the model accuracies.

Keywords

Cite

@article{arxiv.2106.06917,
  title  = {ATRAS: Adversarially Trained Robust Architecture Search},
  author = {Yigit Alparslan and Edward Kim},
  journal= {arXiv preprint arXiv:2106.06917},
  year   = {2021}
}

Comments

9 pages, 2 figures, 2 tables

R2 v1 2026-06-24T03:08:25.414Z