Shared Neural Building Blocks: A Compression Trick That Borrows RNN Logic to Shrink Deep Networks

Shared Neural Building Blocks: A Compression Trick That Borrows RNN Logic to Shrink Deep Networks | BedrockNews