Emergent Low-Rank Training Dynamics in MLPs with Smooth Activations
cs.LG, cs.AI
Submitted: 2026-02-05
Updated: 2026-11-07
Terminology
Sources
- An Overview of Low-Rank Structures in the Training and Adaptation of Large Models
- Gradient Descent Happens in a Tiny Subspace
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- Low Rank Gradients and Where to Find Them
- Understanding Deep Representation Learning via Layerwise Feature Compression and Discrimination
- Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms
- Linearly Separable Features in Shallow Nonlinear Networks: Width Scales Polynomially with Intrinsic Data Dimension
- The Law of Parsimony in Gradient Descent for Learning Deep Linear Networks
Related papers
- 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
- (Mis)generalization of Helpful-Only Fine-tuning