MLR570 · Non-Convex Optimization in Neural Networks

Survey Visual Atlas

Every paper in our survey, each with an interactive visual of its core result, the math behind it, and our study notes. Every surface is a real loss function. Every ball runs a real optimizer.