Research map / Learning theory and algorithms / Kernels and regression theory
Spectra of the Conjugate Kernel and Neural Tangent Kernel for linear-width neural networks
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| train the network rather than keep random initialization | CK and NTK of 3-layer network · single-neuron synthetic target | spectrum shape | Mixed |
| remove leading 10 principal components from CIFAR-10 inputs (ablation) | NTK limit spectrum prediction · CIFAR-10, n=5000 | agreement between observed and theoretical NTK spectrum | Helped |
| remove leading 10 principal components from CIFAR-10 inputs (ablation) | CK limit spectrum prediction · CIFAR-10, n=5000 | agreement between observed and theoretical CK spectrum | No clear effect |
Closest papers by meaning
- Eigen Analysis of Conjugate Kernel and Neural Tangent Kernel (ICML 2025 · similarity 0.93)
- On the Random Conjugate Kernel and Neural Tangent Kernel (ICML 2021 · similarity 0.91)
- Spectral Evolution and Invariance in Linear-width Neural Networks (NeurIPS 2023 · similarity 0.90)
- Understanding the Evolution of the Neural Tangent Kernel at the Edge of Stability (NeurIPS 2025 · similarity 0.87)
- Neural Tangent Kernels Motivate Cross-Covariance Graphs in Neural Networks (ICML 2024 · similarity 0.87)
- Bounds for the smallest eigenvalue of the NTK for arbitrary spherical data of arbitrary dimension (NeurIPS 2024 · similarity 0.87)
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.