AI Paper Atlas

Research map / Adversarial robustness and security / Certified and empirical robustness

Adversarial Weight Perturbation Helps Robust Generalization

Dongxian Wu, Shu-Tao Xia, Yisen Wang

NeurIPS 2020 · Adversarial robustness and security › Certified and empirical robustness · approach: Adversarial training refinements · 204 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Add adversarial weight perturbation to ATAT, WideResNet-34-10 · CIFAR-10, L∞PGD-20 robustness: 56.10 -> 58.14Improved
Add adversarial weight perturbation to TRADESTRADES, WideResNet-34-10 · CIFAR-10, L∞PGD-20 robustness: 56.33 -> 59.27Improved
Add adversarial weight perturbation to MARTMART, WideResNet-34-10 · CIFAR-10, L∞PGD-20 robustness: 58.56 -> 60.68Improved
Add adversarial weight perturbation to RSTRST, WideResNet-28-10 · CIFAR-10, L∞PGD-20 robustness: 62.60 -> 63.73Improved
Add adversarial weight perturbation to RSTRST, WideResNet-28-10 · CIFAR-10, L∞natural accuracy: 89.69 -> 88.25Hurt
Use more than one alternation between input and weight perturbation (ablation)AT-AWP · CIFAR-10test robustnessNo clear effect
Increase weight-perturbation optimization steps K2 from 1 to 5 or 10 (ablation)AT-AWP · CIFAR-10test robustnessNo clear effect
Use L1 or L2/Frobenius norm for weight-perturbation size (ablation)AT-AWP · CIFAR-10test robustnessNo clear effect
Set weight perturbation size too large (γ = 2×10−2) (ablation)AT-AWP · CIFAR-10test robustnessHurt
Replace adversarial weight perturbation with random weight perturbation at γ ≤ 0.3 (ablation)AT-RWP · CIFAR-10test robustnessNo clear effect
Add L1 regularization to AT (ablation)AT · CIFAR-10, L∞best robustness: 52.79 -> 51.95Hurt
Add L2 regularization to AT (ablation)AT · CIFAR-10, L∞best robustness: 52.79 -> 51.60Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy