Skip to content
Free public beta until December 4, 2026: every course is open, no card needed. After that, the first chapter of each course stays free and full access is a one-time $39 ($19 for up to 100 founding members chosen for helpful feedback). Your progress is kept. How the beta works

language-models

Logits and the next-token probability distribution

taught

A model's raw, unnormalized output scores (logits) over the whole vocabulary, and softmax turning them into a probability distribution the model actually samples the next token from.

Taught in

Requires

  • softmax-dl recommended

    deep-learning ch.06/ch.07 derive softmax as a classifier's output layer; this course applies the identical function over a vocabulary-sized output instead of a small label set.

Required by

Nothing in the catalog needs this yet.

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.