Langevin Monte-Carlo Provably Learns Depth Two Neural Nets at Any Size and Data
cs.LG, math.FA, math.PR
Submitted: 2025-03-13
Updated: 2026-11-05
Terminology
Sources
- Concentration of the Langevin Algorithm's Stationary Distribution
- Gradient descent provably escapes saddle points in the training of shallow ReLU networks
- Beating the Perils of Non-Convexity: Guaranteed Training of Neural Networks using Tensor Methods
- Enhanced Convolutional Neural Tangent Kernels
- Adding Gradient Noise Improves Learning for Very Deep Networks
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis