Applied AI

Course map · S1 How LLMs work · Episode 3 of 12

Embeddings

S1 E3 Explainer · 15 slides · 30 September 2026

  1. Slide 1 of 15, Cover.
  2. Slide 2 of 15, Why it matters.
  3. Slide 3 of 15, Mental model.
  4. Slide 4 of 15, Concepts.
  5. Slide 5 of 15, Concepts.
  6. Slide 6 of 15, Data.
  7. Slide 7 of 15, Data.
  8. Slide 8 of 15, Concepts.
  9. Slide 9 of 15, Where it breaks.
  10. Slide 10 of 15, For engineers.
  11. Slide 11 of 15, For engineers.
  12. Slide 12 of 15, In practice.
  13. Slide 13 of 15, For leaders.
  14. Slide 14 of 15, Decision rule.
  15. Slide 15 of 15, Series map.

Lesson notes

Embeddings place text as coordinates on a map of meaning, but similar is not the same as relevant, so treat your embedding model as a versioned dependency.

Decision rule

Treat the embedding model as a versioned dependency: pin it, test it on your own queries, and re-embed when it changes.

Ideas in this lesson

  • Tokens
  • Vectors
  • Cosine similarity
  • Retrieval

Take it further