AI Paper Atlas

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

Adversarial Robustness is at Odds with Lazy Training

Yunjuan Wang, Enayat Ullah, Poorya Mianjy, Raman Arora

NeurIPS 2022 · Adversarial robustness and security › Certified and empirical robustness · approach: Robustness-accuracy trade-off analysis · 1 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
Constrain trained weights to remain in the lazy regimeTwo-layer ReLU networks · Any data distribution on the unit sphererobust error: L_R(\textrm{W};\mathcal{D})\geq 0.9Hurt
Increase the allowed deviation of weights from initialization (ablation)SGD-trained MNIST networks · MNIST 0-vs-1, varying input dimensionminimum perturbation size needed to flip predictionMixed
Increase the perturbation budget beyond 5/\sqrt{d} (ablation)Adversarially trained lazy-regime network · MNIST 0-vs-1, varying input dimensionrobust accuracy: 5/\sqrt{d} -> 0Hurt
Allow the weight radius to exceed the lazy-regime scale (ablation)Adversarially trained network · MNIST 0-vs-1, d = 28^2robust test accuracy: 25/\sqrt{m}Improved
Keep the weight radius below the phase-transition scale (ablation)Adversarially trained network · MNIST 0-vs-1, d = 28^2robust test accuracy: 0Hurt
Project adversarial-training updates to keep weights in the lazy regime (ablation)Adversarially trained MNIST networks · MNIST 0-vs-1test accuracyNo clear effect
Train beyond the lazy regime (ablation)Adversarially trained network · MNIST 0-vs-1robust test accuracyImproved

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