Why adversarial training can hurt robust accuracy

Why adversarial training can hurt robust accuracy

International Conference on Learning Representations (ICLR), 2022

3 March 2022

Julia Hörrmann

ArXiv (abs)PDF HTML

Papers citing "Why adversarial training can hurt robust accuracy"

11 / 11 papers shown

Title
Top-GAP: Integrating Size Priors in CNNs for more Interpretability, Robustness, and Bias Mitigation Lars Nieradzik Henrike Stephani Janis Keuper FAtt AAML 162 1 0 07 Sep 2024
Approximate Nullspace Augmented Finetuning for Robust Vision Transformers Haoyang Liu Aditya Singh Yijiang Li Haohan Wang AAML ViT 238 1 0 15 Mar 2024
Specification Overfitting in Artificial IntelligenceArtificial Intelligence Review (Artif Intell Rev), 2024 Benjamin Roth Pedro Henrique Luz de Araujo Yuxi Xia Saskia Kaltenbrunner Christoph Korab 461 8 0 13 Mar 2024
Adversarially Robust Feature Learning for Breast Cancer Diagnosis Degan Hao Dooman Arefan M. Zuley Wendie Berg Shandong Wu OOD MedIm 128 4 0 13 Feb 2024
Accuracy of TextFooler black box adversarial attacks on 01 loss sign activation neural network ensemble Yunzhe Xue Usman Roshan AAML 126 0 0 12 Feb 2024
Improving the Robustness of Transformer-based Large Language Models with Dynamic AttentionNetwork and Distributed System Security Symposium (NDSS), 2023 Lujia Shen Yuwen Pu R. Beyah Changjiang Li Xuhong Zhang Chunpeng Ge Ting Wang AAML 119 8 0 29 Nov 2023
Theoretical Analysis of Robust Overfitting for Wide DNNs: An NTK ApproachIEEE Transactions on Information Theory (IEEE Trans. Inf. Theory), 2023 Shaopeng Fu Haiyan Zhao AAML 223 3 0 09 Oct 2023
Mitigating Adversarial Attacks in Federated Learning with Trusted Execution EnvironmentsIEEE International Conference on Distributed Computing Systems (ICDCS), 2023 Simon Queyrut V. Schiavoni Pascal Felber AAML FedML 132 13 0 13 Sep 2023
Pelta: Shielding Transformers to Mitigate Evasion Attacks in Federated Learning Simon Queyrut Yérom-David Bromberg V. Schiavoni FedML AAML 126 1 0 08 Aug 2023
How robust accuracy suffers from certified training with convex relaxations Piersilvio De Bartolomeis Jacob Clarysse Amartya Sanyal Fanny Yang AAML 126 2 0 12 Jun 2023
Devil is in Channels: Contrastive Single Domain Generalization for Medical Image SegmentationInternational Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI), 2023 Shishuai Hu Zehui Liao Yong-quan Xia MedIm 239 44 0 08 Jun 2023