Deep Learning Theory, Neural Network Pruning

How Much Pre-training Is Enough to Discover a Good Subnetwork?

Strong Lottery Ticket Hypothesis with $\epilon$–perturbation