Research map / Optimization / Implicit bias and generalization
Transition to Linearity of Wide Neural Networks is an Emerging Property of Assembling Weak Models
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 |
|---|---|---|---|
| Increase hidden-layer width | Fully connected neural network · CIFAR-10 subset (20 airplane or bird images) | average absolute gradient cosine | Helped |
| Increase bottleneck-layer width | Bottleneck neural network · CIFAR-10 subset (20 airplane or bird images) | average absolute gradient cosine | Helped |
| Assemble more independent, non-dominating sub-models with the stated scaling | Assembly model · O(1)-neighborhood of initialization | Taylor quadratic-term magnitude: O(log(m)/√m) | Helped |
| Increase network width | Wide fully connected neural network · O(1)-neighborhood of initialization | Taylor quadratic-term magnitude | Helped |
Closest papers by meaning
- On the linearity of large non-linear models: when and why the tangent kernel is constant (NeurIPS 2020 · similarity 0.92)
- Wide Neural Networks Trained with Weight Decay Provably Exhibit Neural Collapse (ICLR 2025 · similarity 0.87)
- Deep Linear Network Training Dynamics from Random Initialization: Data, Width, Depth, and Hyperparameter Transfer (ICML 2025 · similarity 0.86)
- Three Mechanisms of Feature Learning in a Linear Network (ICLR 2025 · similarity 0.86)
- A Dynamical Model of Neural Scaling Laws (ICML 2024 · similarity 0.86)
- Stable Minima of ReLU Neural Networks Suffer from the Curse of Dimensionality: The Neural Shattering Phenomenon (NeurIPS 2025 · similarity 0.86)
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.