Research map / Optimization / Implicit bias and generalization
Hausdorff Dimension, Heavy Tails, and Generalization in 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 |
|---|---|---|---|
| Decrease the stable-process tail index α | Stable-process training algorithm · Synthetic logistic regression | generalization error | Helped |
| Use estimated βS to characterize trajectory complexity | VGG networks trained with SGD · CIFAR-10 | generalization error | Mixed |
| Use η/B as the generalization metric | SGD-trained VGG networks · CIFAR-10 | correlation with generalization error | Mixed |
| Use the heavy-tailed random-matrix metric | SGD-trained VGG networks · CIFAR-10 | correlation with generalization error | Mixed |
| Increase network depth | SGD-trained VGG networks · CIFAR-10 | generalization error | Mixed |
| Vary the SGD step size η and batch size B | SGD-trained VGG networks · CIFAR-10 | estimated βS | Mixed |
Closest papers by meaning
- The Heavy-Tail Phenomenon in SGD (ICML 2021 · similarity 0.93)
- Emergence of heavy tails in homogenized stochastic gradient descent (NeurIPS 2024 · similarity 0.91)
- A Tail-Index Analysis of Stochastic Gradient Noise in Deep Neural Networks (ICML 2019 · similarity 0.91)
- Generalization Bounds for Heavy-Tailed SDEs through the Fractional Fokker-Planck Equation (ICML 2024 · similarity 0.89)
- Error Analysis Affected by Heavy-Tailed Gradients for Non-Convex Pairwise Stochastic Gradient Descent (AAAI 2025 · similarity 0.88)
- The Global Convergence Time of Stochastic Gradient Descent in Non-Convex Landscapes: Sharp Estimates via Large Deviations (ICML 2025 · similarity 0.88)
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.