Research map / Optimization / Implicit bias and generalization
Gradient Descent in Neural Networks as Sequential Learning in Reproducing Kernel Banach Space
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 |
|---|---|---|---|
| Replace first-order Taylor (NTK) approximation with exact power-series representation of the network in a finite neighbourhood of initial weights, expressed as inner product of data and weight-step feature maps | fully connected neural network trained by gradient descent · theoretical analysis (no experiments) | not reported | No clear effect |
| Regularized sequential learning in RKBS with appropriate scale factors, shadow weights and trade-off lambda reproduces gradient descent weight-step exactly | gradient descent weight-step in neural networks · theoretical equivalence (Theorem 1) | equality of weight-step | Helped |
Closest papers by meaning
- Neural Tangent Kernel: Convergence and Generalization in Neural Networks (NeurIPS 2018 · similarity 0.92)
- On the Impacts of the Random Initialization in the Neural Tangent Kernel Theory (NeurIPS 2024 · similarity 0.91)
- Optimal Rates for Averaged Stochastic Gradient Descent under Neural Tangent Kernel Regime (ICLR 2021 · similarity 0.90)
- Scaling Neural Tangent Kernels via Sketching and Random Features (NeurIPS 2021 · similarity 0.89)
- Convergence Analysis of Newton's Method for Neural Networks in the Overparameterized Limit (NeurIPS 2026 · similarity 0.89)
- A generalized neural tangent kernel for surrogate gradient learning (NeurIPS 2024 · similarity 0.89)
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.