Overview
Softmax is used for multi-class classification. It takes the raw scores (logits) from the network and converts them into probabilities for each possible class.
How it Works
It exponentiates each score and then divides by the sum of all exponentiated scores. This ensures the largest score gets the highest probability.
Application
If a model is classifying an image as a 'cat,' 'dog,' or 'bird,' Softmax will provide the probability for each, such as [0.1, 0.8, 0.1].