deep-learning
The softmax function (raw scores to probabilities)
taught
Taught in
-
From scores to probabilities: the softmax
Deep Learning / Neural Networks
-
Softmax and cross-entropy recalled
Wiring Language Models
Requires
- the-sigmoid-function required
Required by
- Categorical cross-entropy required
- Logits and the next-token probability distribution recommended
- Reductions: .sum(), .mean(), .argmax() helpful
Local knowledge map
Local knowledge map
Click to expand
Solid arrows are required prerequisites; dashed are recommended or helpful. Lower rows are more foundational. Click any box to jump to that concept.