activity
20202026
most citedOn the training of sparse and dense deep neural networks: less parameters, same performance

13 citations · 35 across the 16 of their papers we have counts for

collaborators
Showing 2024Show all

4 papers · 1 filter