Hacker News new | ask | show | jobs
by jmalicki 49 days ago
The cross entropy loss function is softmax. They are one and the same.
1 comments

They’re not. Cross entropy loss is E[-log q] where q is a probability. You could convert the model outputs x into probabilities using some other function like q = 1/Z x^2, and compute cross entropy loss just fine.
Behold the actual definition of cross entropy: https://en.wikipedia.org/wiki/Cross-entropy

It's true that the PyTorch API conflates cross entropy and softmax, but they are separate concepts.