Bach

Reading group. “Gradient descent for wide two-layer neural networks”. [Blog post]