Phase diagram for two-layer ReLU neural networks at infinite-width limit
From MaRDI portal
Publication:4998974
Recommendations
- Infinite-width limit of deep linear neural networks
- Disentangling feature and lazy training in deep neural networks
- Wide neural networks of any depth evolve as linear models under gradient descent *
- Phase diagram of stochastic gradient descent in high-dimensional two-layer neural networks
- Mean Field Analysis of Deep Neural Networks
Cites work
- A comparative analysis of optimization and generalization properties of two-layer neural network and random feature models under gradient descent dynamics
- A mean field view of the landscape of two-layer neural networks
- Disentangling feature and lazy training in deep neural networks
- High-dimensional probability. An introduction with applications in data science
- Mean field analysis of neural networks: a central limit theorem
Cited in
(19)- Weight-decay induced phase transitions in multilayer neural networks
- Adaptive estimation of nonparametric functionals
- On the Explainability of Graph Convolutional Network With GCN Tangent Kernel
- Loss jump during loss switch in solving PDEs with neural networks
- Homotopy relaxation training algorithms for infinite-width two-layer ReLU neural networks
- Optimistic estimate uncovers the potential of nonlinear models
- Phase diagram of initial condensation for two-layer neural networks
- Local linear recovery guarantee of deep neural networks at overparameterization
- Understanding the initial condensation of convolutional neural networks
- Overview frequency principle/spectral bias in deep learning
- Embedding principle: a hierarchical structure of loss landscape of deep neural networks
- Modeling structured data learning with restricted Boltzmann machines in the teacher-student setting
- Overlapping Schwarz preconditioners for randomized neural networks with domain decomposition
- On understanding and overcoming spectral biases of deep neural network learning methods for solving PDEs
- Geometric structure of shallow neural networks and constructive \(\mathcal{L}^2\) cost minimization
- Modern and emerging phenomena in machine learning. Abstracts from the workshop held March 8--13, 2026
- Towards understanding gradient flow dynamics of homogeneous neural networks beyond the origin
- Anchor function: a type of benchmark functions for studying language models
- Singular parameters and missing limits in neural PDE solvers
This page was built for publication: Phase diagram for two-layer ReLU neural networks at infinite-width limit
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4998974)
