Turns a list of scores into probabilities with a total of 1.
Softmax is like cutting one pizza for a hungry table. The loudest kid gets the biggest slice.
It turns raw scores into fair-looking chances. You see it when AI picks a label or the next word.
Token
Softmax turns each candidate Token score into a probability.
Temperature
Temperature can flatten or sharpen Softmax, so output feels more random or more focused.
LLM
An LLM often uses Softmax to choose the next word.
Backpropagation
Backprop sends the error signal back through Softmax during training.