# W1: Token, cost, and temperature worksheet

How Large Language Models Work: Foundations for Choosing and Applying LLMs at Work (FAST-14) · Lab 1: Tokens, probabilities, and temperature

Name: ______________________  Date: ____________

| Item | Your result |
| --- | --- |
| FAQ words / tokens / tokens per word | |
| String that split into the most tokens, and why you think so | |
| Tokens per request (FAQ + instructions + question) | |
| Cost per question (show the formula) | |
| Monthly cost, 3,000 questions; with caching | |
| Top 5 next tokens for "every ..." and their probabilities | |
| Temperature 0 vs 1.0: how much did answers vary for welcome vs check-out? | |
| Setting you would choose for the FAQ bot, with one-sentence reason | |

**Completion criteria**

- Token counts are recorded and the cost is calculated with the formula shown.
- The answer separates input tokens from output tokens.
- The temperature choice is justified by the type of task (factual versus creative).