no code implementations • 1 Jan 2021 • Ido Bronstein, Alon Brutzkus, Amir Globerson
Here we focus on this setting where the functions are learned by a convex neural network and gradient descent.
Inductive Bias