Matlab Code of paper "Implicit Regularization in ReLU Networks with the Square Loss"
This paper is by Gal Vardi, Ohad Shamir. You can find the paper on https://arxiv.org/abs/2012.05156
In the paper they used L2 loss, however I added L1 loss, log(cosh) loss and Huber loss to my code.
File plot_Implicit_Regularization.m contains the function. File Test.m is an example that you could see how it works with different loss functions. Enjoy ;)