Toggle navigation
chex
Archive
Categories
Pages
Tags
Submit
Archive
2014
August
August 7, 2014
»
Why do we rotate weights when computing the gradients in a convolution layer of a convolution network?