Summary  
This chapter covers techniques for guiding AI output by presenting examples of the desired result (few-shot prompting) and by encouraging step-by-step reasoning for complex tasks (chain-of-thought prompting).

General domain of usage  
Machine translation

Example-based prompting involves showing the AI one or more examples of the desired output. This is called **few-shot prompting**. 

**Few-shot prompting**: supplying a few examples in the prompt to demonstrate the desired output.

Definition

For complex or multi-step tasks, **chain-of-thought prompting** encourages the AI to reason step by step.



**Chain-of-thought prompting**: encouraging the AI to reason through a problem step by step in its response.

#### Example comparison

- **Few-shot:** `Translate the following sentences to French. Example: "Good morning." → "Bonjour." Now translate: "How are you?"`;
- **Chain-of-thought:** `Solve this math problem step by step: If a train travels 60 miles in 1.5 hours, what is its average speed?`.

Use **few-shot prompting** when you want the AI to mimic a specific style or format. Use **chain-of-thought prompting** for tasks that require logical steps or explanations.

Too many examples can overwhelm the AI or exceed its context window. Choose **concise** and **relevant** examples.

Note

When should you use chain-of-thought prompting?

A beginner-friendly, fully theoretical course introducing the core concepts and practical strategies of prompt engineering for large language models (LLMs). Learn how to structure, refine, and optimize prompts to achieve better AI-generated results.

Explore the essential building blocks of prompt engineering, including what prompts are, why they matter, and the key principles for effective communication with LLMs.

Practical prompting techniques for controlling AI behavior and improving response quality.

Learn how to refine, evaluate, and improve prompts for consistency and real-world use cases.

Example-Based and Step-by-Step Prompting

Example comparison