Improving Lipschitz-Constrained Neural Networks by Learning Activation Functions

Open in new window