Do Perceptually Aligned Gradients Imply Adversarial Robustness?

2022/07/22 by Roy Ganz, Ganz, Roy, Bahjat Kawar +3 · 5 citations
Computer Science · #Adversarial Robustness in Machine Learning #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences

paper · pdf · doi:10.48550/arxiv.2207.11378

openalex publication_date 2022/07/22 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Adversarially robust classifiers possess a trait that non-robust models do not -- Perceptually Aligned Gradients (PAG). Their gradients with respect to the input align well with human perception. Several works have identified PAG as a byproduct of robust training, but none have considered it as a standalone phenomenon nor studied its own implications. In this work, we focus on this trait and test whether Perceptually Aligned Gradients imply Robustness. To this end, we develop a novel objective to directly promote PAG in training classifiers and examine whether models with such gradients are more robust to adversarial attacks. Extensive experiments on multiple datasets and architectures validate that models with aligned gradients exhibit significant robustness, exposing the surprising bidirectional connection between PAG and robustness. Lastly, we show that better gradient alignment leads to increased robustness and harness this observation to boost the robustness of existing adversarial training techniques.

Cited by

• DeepDefense: Layer-Wise Gradient-Feature Alignment for Building Robust Neural Networks
• Adversarial Attacks Leverage Interference Between Features in Superposition
• Inverse Language Modeling towards Robust and Grounded LLMs
• Adversarial Examples Are Not Bugs, They Are Superposition
• LeakyCLIP: Extracting Training Data from CLIP

Related

• Are Perceptually-Aligned Gradients a General Property of Robust Classifiers?
• On the Benefits of Models with Perceptually-Aligned Gradients
• Which Models have Perceptually-Aligned Gradients? An Explanation via Off-Manifold Robustness
• Bridging Adversarial Robustness and Gradient Interpretability
• What it Thinks is Important is Important: Robustness Transfers through Input Gradients