"So the inputs to this module on the backward pass are X-IN, which we can write as the gradient vector .X-IN Let’s compute the backward pass output: " X-IN Should be X_outs
"So the inputs to this module on the backward pass are
X-IN, which we can write as the gradient vector
.X-IN Let’s compute the backward pass output: "
X-IN Should be X_outs