An interactive tutorial
How language models
actually work, from the ground up.
A first-principles path through tokens, embeddings, attention, and the machinery behind LLMs, built up step by step, with worked examples and runnable code that make it stick.
10 lessons across 1 chapter · ~48 min · free to read
Highlight & note
Mark up the text. Your highlights and notes sync across devices.
Spaced review
Key ideas resurface on an FSRS schedule so they stay learned.
Exercises
Predict tokenizer output, trace algorithms, and check your understanding with instant feedback.
Read freely, no account needed
Every chapter is open to everyone. Signed out, the site stores nothing on a server and tracks nothing, so you lose no content by staying logged out.
Sign in to unlock everything else
Reading is free. Everything beyond it needs a free account: progress tracking, highlights and notes, dark and sepia themes, full-screen focus mode, and spaced review, all synced across your devices.
✓ You're signed in. Your personal layer is on.
Your data syncs across devices and stays private to your account.
Your privacy, plainly.
Reading never requires an account and is never tracked. If you create an account, we store only your email and your own learning data: progress, highlights, and notes. We never sell your data. We never show ads. Delete your account anytime and it is gone for good.