Foundational Intelligence
Build the mental model first. This course explains what an LLM is, how tokens work, why words become numbers, and how attention lets a model focus.
Progress loads on this device after the page opens.
A beginner-friendly introduction to the mechanics behind modern language models, written to remove jargon before it creates confusion.
Move through the course in order.
Phase 1. Core mechanics
Understand the basic loop that turns text into predictions and predictions into full answers.
- 1What Is an LLM?
Large Language Models power tools like ChatGPT, but the mechanism behind them is far simpler than most people think. This lesson explains the core idea clearly: how predicting the next token at massive scale produces systems that appear intelligent. The model does not understand meaning the way humans do. It has simply learned which pieces of language usually appear together.
18 min · Core lessonNext - 2How AI Sees Your Words
Computers cannot understand words the way humans do. Before an AI system can work with language, it must first break text into small pieces called tokens. In this lesson we explore how language is split into these pieces and how LLMs generate text by predicting them one step at a time.
15 min · Core lessonOpen - 3How Words Become Numbers
Computers cannot learn from text unless it is converted into numbers. In this lesson we explore how tokens are transformed into numerical representations called embeddings, allowing neural networks to learn patterns in language.
15 min · Core lessonOpen - 4How AI Decides What to Pay Attention To
When a model reads a sentence, not every word matters equally. In this lesson we learn how transformer models use attention to focus on the most relevant parts of the text when predicting the next token.
15 min · Core lessonOpen