Discover few-shot prompting

In prompt engineering, the word “shot” ​is often used as a synonym for the word “example.” ​There are different names for prompting techniques ​based on the number of examples given to the LLM.

Zero-shot prompting is a technique ​that provides no examples in a prompt

  • Because examples aren’t included in zero-shot prompts, ​the model is expected to perform the task ​based only on its training data, ​and the task description included in the prompt. ​
  • Zero-shot prompting is most likely to be effective ​when you’re seeking simple and direct responses. ​
  • Zero-shot prompting may not be effective ​for tasks that require the LLM ​to respond in a more specific, nuanced way

​while one-shot prompting provides one example, ​and few-shot prompting is a technique ​that provides two or more examples in a prompt

  •  ​Few shot prompting can improve in LLMs performance ​by providing additional context and examples in your prompt. ​
  • These additional examples can help clarify ​the desired format, phrasing, or general pattern. ​