Repaso
EspaƱol

What is few-shot prompting?

Options

  1. Including a few examples of input and the desired output inside the prompt (Correct option)
  2. Calling the model a few times and keeping the best answer
  3. Fine-tuning the model on a handful of examples
  4. Using a small model for short tasks

Answer

Including a few examples of input and the desired output inside the prompt

Why

In few-shot prompting you show the model two to five examples of the task in the prompt itself, and it follows the pattern for the new input.

No weights change: the examples live only in that request and disappear after it.

Fine-tuning is different: it trains the model on examples ahead of time and changes it.

Remember

Few-shot = examples inside the prompt.

Key terms

Practice this question in Repaso