Module 1
How LLMs Work
Build a correct mental model of tokens, embeddings, attention, and next-token generation.
~90 minutes
Learning path
- 1How a chat app talks to an AI model~8 min · Guest free
- 2AI isn’t just ChatGPT~14 min · Guest free
- 3What Is a Model?~12 min · Guest free
- 4How Does a Model Learn?~12 min
- 5Training vs Inference~10 min
- 6What’s inside the model?~10 min
- 7Tokens: how models chop text~12 min
- 8Context windows: a finite budget~11 min
- 9Embeddings: meaning as numbers~12 min
- 10Why Transformers?~11 min
- 11Attention: what matters right now~11 min
- 12Query, Key, Value~11 min
- 13Self-attention: looking inside the same text~12 min
- 14A Transformer block (one Lego brick)~10 min
- 15Translate wiring vs chat wiring~9 min
- 16What is an LLM?~10 min
- 17Next token: one piece at a time~10 min
- 18From raw scores to percentages~9 min
- 19Sampling: how we pick the next token~11 min
- 20Why can an LLM follow instructions?~10 min
Milestone
Build a Mental Model of an LLM
You have the pieces. Assemble one path that turns raw text into a generated next token. When it works, we change one constraint and ask you to diagnose the failure like an engineer.