Fetching the paper…

Overparameterized Nonlinear Learning: Gradient Descent Takes the Shortest Path? · Around