A free course · English & Português
How language models actually work.
From “what is a parameter?” to streaming a 2.8-trillion-parameter model off a disk on an 8 GB laptop. Every claim cited, every lab computing real numbers, nothing behind a login.
79 lessons available now
How it works
- ConceptThe actual mechanism, no hand-waving.
- AnalogyThe same idea in plain language.
- LabReal numbers, computed live.
- Teach-backYou explain it, then compare with a model answer.
- QuizWith the reason the right answer is right.
A lesson counts as complete only when your teach-back has substance and the quiz is right. Clicking “reveal” is not learning. Everything stays in your browser.
Tracks
- foundationsOrientation & FoundationsBuild an accurate mental model of language models and the mathematics needed for everything that follows.8 lessons
- foundationsFrom Text to TensorsSee how text becomes tokens, vectors, and numerical structures a neural network can process.7 lessons
- foundationsNeural Network FundamentalsLearn how neural networks represent functions, measure error, propagate gradients, and improve through optimization.9 lessons
- coreSequence ModelsFollow the path from recurrent state to encoder-decoder models and attention, the immediate ancestors of the Transformer.5 lessons
- coreThe TransformerAssemble the architecture behind modern LLMs, beginning with self-attention and its query, key, and value operations.14 lessons
- advancedPretraining at ScaleDesign the objectives, data, optimization, parallelism, and cost model behind a serious language-model pretraining run.13 lessons
- advancedPost-training & AlignmentTurn pretrained predictors into useful assistants through instruction tuning, preference learning, reinforcement learning, and distillation.11 lessons
- advancedInference & EfficiencyUnderstand how trained models generate text and how decoding choices trade off diversity, coherence, latency, and cost.12 lessons
More tracks are being written. What is published is finished — there are no half-written lessons.