# Adversarially Robust Learning via Entropic Regularization

arXiv: https://arxiv.org/abs/2008.12338 · HTML: https://mlsanity.com/paper/2008.12338/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 27 Aug 2020.

## Built on by (most cited first, 9 of 9)

- [Sparsity Winning Twice: Better Robust Generalization from More Efficient Training](https://mlsanity.com/paper/2202.09844/) (2022)
- [Randomized Adversarial Training via Taylor Expansion](https://mlsanity.com/paper/2303.10653/) (2023)
- [Adversarial Token Attacks on Vision Transformers](https://mlsanity.com/paper/2110.04337/) (2021)
- [AERO: Entropy-Guided Framework for Private LLM Inference](https://mlsanity.com/paper/2410.13060/) (2024)
- [SafeAMC: Adversarial training for robust modulation recognition models](https://mlsanity.com/paper/2105.13746/) (2021)
- [ALMOST: Adversarial Learning to Mitigate Oracle-less ML Attacks via Synthesis Tuning](https://mlsanity.com/paper/2303.03372/) (2023)
- [Towards Unbiased Exploration in Partial Label Learning](https://mlsanity.com/paper/2307.00465/) (2023)
- [ALTER: Asymmetric LoRA for Token-Entropy-Guided Unlearning of LLMs](https://mlsanity.com/paper/2603.01792/) (2026)
- [Towards Generalized Entropic Sparsification for Convolutional Neural Networks](https://mlsanity.com/paper/2404.04734/) (2024)

## Abstract (arXiv)

In this paper we propose a new family of algorithms, ATENT, for training adversarially robust deep neural networks. We formulate a new loss function that is equipped with an additional entropic regularization. Our loss function considers the contribution of adversarial samples that are drawn from a specially designed distribution in the data space that assigns high probability to points with high loss and in the immediate neighborhood of training samples. Our proposed algorithms optimize this loss to seek adversarially robust valleys of the loss landscape. Our approach achieves competitive (or better) performance in terms of robust classification accuracy as compared to several state-of-the-art robust learning approaches on benchmark datasets such as MNIST and CIFAR-10.