Khalid ShamsPrincipal Solutions Architect

Learn AI · Guided path 3 of 6

What a language model does

When you finish: You can explain tokens, context and attention well enough to know why a model forgets, drifts and sometimes makes things up.

About 20 minutes · 10 stops, 9 of them ideas from the map · progress is kept on this device ·

Start at stop 1 All paths

The stops, in order

Each stop is one page. Read it, picture it, mark it, move on.

Ideas open on their own page with three depths. Atlas stops open the interactive lesson at the right step and count as done when you follow the run to its end. Sections count when you have read them.

  1. 1TokensThe small chunks of text an AI actually reads, often word pieces rather than whole words.idea
  2. 2EmbeddingsTurning a word or idea into a list of numbers so that similar meanings end up close together.idea
  3. 3AttentionA way for a model to decide which earlier words matter most for understanding the current one.idea
  4. 4TransformerThe 2017 design that nearly all modern language AI is built on.idea
  5. 5Large Language ModelA program that writes by predicting the next word, over and over, very well.idea
  6. 6Context WindowHow much text an AI can keep in mind at once.idea
  7. 7PromptingWhat you type to an AI. Clear context and a clear ask get clearer answers.idea
  8. 8System PromptThe standing instructions a model is given before the conversation starts: who it is, what it may do, how to answer.idea
  9. 9HallucinationWhen an AI states something false with total confidence.idea
  10. 10Watch a model propose, not actStep 5 of the flagship lesson, then follow the run to the end.atlas lesson

← Path 2 Path 4: What makes it an agent →