AI Paper Atlas

Research map / Optimization / Implicit bias and generalization

When Do Neural Networks Outperform Kernel Methods?

Behrooz Ghorbani, Song Mei, Theodor Misiakiewicz, Andrea Montanari

NeurIPS 2020 · Optimization › Implicit bias and generalization · approach: Neural tangent kernel analysis · 70 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
Randomize 1.5% of low-frequency componentsMyrtle-5 CNN and Myrtle-5 CNTK KRR · CIFAR-10classification accuracy: 87.7% -> 84.7%; 77.6% -> 58.2%Mixed
Increase high-frequency noise in the training and test imagesNeural networks and RKHS methods · FMNISTclassification accuracyMixed
Increase high-frequency noise in the training and test imagesNeural networks and RKHS methods · CIFAR-2classification accuracy gap between NN and NT KRR: 0.6% at tau = 0 -> 4.5% at tau = 3Improved
Replace low-frequency image components with Gaussian noiseNeural networks and RKHS methods · FMNISTclassification accuracyMixed
Replace low-frequency image components with Gaussian noiseMyrtle-5 CNN and CNTK KRR · CIFAR-10classification accuracyMixed

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