Pruning Neural Networks at Initialization: Why are We Missing the Mark?

Open in new window