softmax
/ˈsɔːftmæks/
a function converting raw scores into a probability distribution
softmax in a sentence
“Softmax ensures the output probabilities sum to one.”
Origin of softmax
Soft (smooth) + max (maximum); mathematical term
Related Words
transformer
a neural network architecture using self-attention for sequence processing
embedding
a dense vector representation of discrete items like words
attention mechanism
a technique allowing models to focus on relevant parts of input
latent space
a compressed representation where similar items are close together
token
a unit of text (word, subword, or character) processed by a model
weight
a learnable parameter that determines connection strength in a network