DYAD: A Descriptive Yet Abjuring Density efficient approximation to linear neural network layers

Open in new window