Are ResNets Provably Better than Linear Predictors?

Open in new window