Non Robust Feature

Non-robust features, easily manipulated aspects of data that influence model predictions but lack genuine representational power, are a central concern in machine learning robustness research. Current work focuses on identifying and mitigating the influence of these features through techniques like input density smoothing and information bottleneck methods, often within the context of adversarial training and various deep learning architectures. Understanding and controlling the impact of non-robust features is crucial for building trustworthy and reliable machine learning systems, particularly in safety-critical applications where model predictions must be resistant to manipulation.

Papers

July 8, 2024

Non-Robust Features are Not Always Useful in One-Class Classification
Matthew Lau, Haoran Wang, Alec Helbling, Matthew Hul, ShengYun Peng, Martin Andreoni, Willian T. Lunardi, Wenke Lee
Adversarial Attack Adversarial Example One Class Classification Lightweight Classifier Non Robust Feature

July 5, 2024

Regulating Model Reliance on Non-Robust Features by Smoothing Input Marginal Density
Peiyu Yang, Naveed Akhtar, Mubarak Shah, Ajmal Mian
Feature Attribution Robustness Problem Model Trustworthiness Non Robust Feature

April 16, 2024

Do Counterfactual Examples Complicate Adversarial Training?
Eric Yeats, Cameron Darwin, Eduardo Ortega, Frank Liu, Hai Li
Robust Model Robust Classifier Counterfactual Example Robustness Trade Counterfactual Generative Non Robust Feature

October 29, 2023

Adversarial Examples Are Not Real Features
Ang Li, Yifei Wang, Yiwen Guo, Yisen Wang
Native Robustness Adversarial Example Adversarial Vulnerability Non Robust Feature

November 21, 2022

Addressing Mistake Severity in Neural Networks with Semantic Knowledge
Natalie Abreu, Nathan Vaska, Victoria Helus
Neural Network Deep Neural Network Adversarial Attack Native Robustness Semantic Knowledge Error Mitigation Standard Adversarial Training Non Robust Feature

November 20, 2022

Feature Weaken: Vicinal Data Augmentation for Classification
Songhao Jiang, Yan Chu, Tianxing Ma, Tianning Zang
Classification Code Adversarial Sample Novel Training Feature Mixing Non Robust Feature

June 16, 2022

Catastrophic overfitting can be induced with discriminative non-robust features
Guillermo Ortiz-Jiménez, Pau de Jorge, Amartya Sanyal, Adel Bibi, Puneet K. Dokania, Pascal Frossard, Gregory Rogéz, Philip H. S. Torr
Native Robustness Adversarial Training Catastrophic Overfitting Robust Neural Network Robustness Problem Non Robust Feature

June 11, 2022

Improving the Adversarial Robustness of NLP Models by Information Bottleneck
Cenyuan Zhang, Xiang Zhou, Yixin Wan, Xiaoqing Zheng, Kai-Wei Chang, Cho-Jui Hsieh
Adversarial Example Adversarial Robustness Information Bottleneck NLP Model Robust Feature Non Robust Feature

April 26, 2022

On Fragile Features and Batch Normalization in Adversarial Training
Nils Philipp Walter, David Stutz, Bernt Schiele
Adversarial Example Adversarial Training Adversarial Robustness Batch Normalization Non Robust Feature

April 6, 2022

Distilling Robust and Non-Robust Features in Adversarial Examples by Information Bottleneck
Junho Kim, Byung-Kwan Lee, Yong Man Ro
Adversarial Example Information Bottleneck Model Robustness Adversarial Prediction Non Robust Feature

January 31, 2022

Can Adversarial Training Be Manipulated By Non-Robust Features?
Lue Tao, Lei Feng, Hongxin Wei, Jinfeng Yi, Sheng-Jun Huang, Songcan Chen
Adversarial Training Availability Attack Non Robust Feature Diffusion Attack Test Time Adversarial