lchizat/2020-implicit-bias-wide-2NN

Stars
7
Rank 2,294,772 (Top 46 %)
Language
Julia
License
MIT License
Created almost 5 years ago
Updated almost 5 years ago

lchizat/2020-implicit-bias-wide-2NN

lchizat

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

Code for the paper: "Implicit Bias of Gradient Descent for Wide Two-layer Neural Networks Trained with the Logistic Loss" (Chizat and Bach)

optimal-transport

Optimal transport and generalizations

Jupyter Notebook

2020_implicitbias_blog

Code for the figure of the "implicit bias" blog post

Jupyter Notebook

2019-sparse-optim-measures

Source code for the paper "Sparse Optimization on Measures with Over-parameterized Gradient Descent"

illustrations

2021-measures-PGM

Code for the paper "Convergence Rates of Gradient Methods for Convex Optimization in the Space of Measures"

Jupyter Notebook

2022-wide-linear-NN

Dynamics of infinitely wide linear neural nets

Jupyter Notebook

2022-mean-field-langevin-rate

Code for the paper "Mean-Field Langevin Dynamics : exponential convergence and annealing"

2021-exp-ICM

Jupyter Notebook