Better Depth-Width Trade-offs for Neural Networks through the lens of Dynamical Systems

Open in new window