A learning plan built around you
Understand and use large language models, with a plan built around your goal, your time and your work.
Early preview. Free, no sign-up. Full personalised plans are still being built.
From where you are to where you want to be
Most courses start at the beginning. This one starts with you.
Take the diagnostic
Four questions, and for each you say how sure you are. You get an honest picture of what you know, what you half know, and where the gaps are.
Get a plan that fits
A dated plan built around your goal, your time each week and any deadline, with a calendar export. If you fall behind, it re-paces itself.
Learn it so it stays learned
Short lessons with practice, then review that brings an idea back just before you would forget it. That recall is what makes it last.
The fastest way to understand a model is to poke at one
The LLM Lab is free, with no sign-up. Small interactive tools let you watch the ideas from the lessons working.
A language model never knows the answer. It only predicts the next token.
- Tokenizer playgroundSee how a real tokenizer splits any text.
- Temperature labReshape next-word probabilities, then sample.
- Context window labWatch the oldest messages fall out of view.
- Embeddings mapMeaning, drawn as a map of real words.
- RAG builderChoose what to retrieve, see the answer change.
- Spot the hallucinationFind the fabricated details among true ones.
The companion to the book
Ascend goes with How LLMs Actually Work, the plain-English book on how ChatGPT, Claude and Gemini work, from tokens to transformers and from hardware to hallucination.
About the bookFind out where you stand
Four questions, about two minutes, no sign-up.