Awesome Testing
Request training offer
Guided course + reference

How LLMs Work materials

Every lesson moves through a concise mechanism, experiment, written evidence brief, and checkpoint. Drafts stay in your browser and can be exported as Markdown; the complete theory chapter remains available afterward as optional reference.

Prior knowledge
Basic vectors and probability help. The machine-learning course is useful, but not required.
Study time
3–4 hours guided · about 8 hours of optional deep dives
By the end
Trace a prompt through tokenization, representations, attention, transformer blocks, logits, sampling, and cached generation.

10

lesson experiments

10

exportable worksheets

10

reference chapters

10

knowledge checks

Start guided course

Canonical learner path

Course chapters

Each route keeps the mandatory learning loop concise, saves its worksheet independently, and loads the long-form chapter only when the learner opens the deep dive.

  1. 01The destinationWhat should follow “The animal did not cross the street because it was too …”?
  2. 02Text becomes tokensWhat exact sequence reaches the neural network?
  3. 03IDs become vectorsHow can a categorical ID become something a neural network can transform?
  4. 04Two jobs per blockWhat changes a token state inside one transformer block?
  5. 05Attention gathers contextWhich earlier positions should “too” consult before the model predicts the next token?
  6. 06Updates accumulateWhere do attention and MLP updates go after each sublayer?
  7. 07The model makes a choiceHow does the final state become “tired” and its alternatives?
  8. 08Generation repeatsWhat can the model reuse when the chosen token extends the sequence?
  9. 09The error travels backwardWhat happens when the observed next token is “tired” but its probability is low?
  10. 10Explain and evaluateIf “animal” becomes “robot,” what changes—and how would you test whether the system still works?