Q: A graphic artist writes a prompt asking a large language model to summarize the latest trends in the design industry. Within the prompt, they provide three examples to demonstrate the context, phrasing, and intended format of the output. What does this scenario describe?
or
Q: An extensive language model receives a suggestion from a graphic designer summarizing the most recent developments in the design sector. They provide three samples in the prompt to show the desired format, language, and context of the output. What is described in this scenario?
- One-shot prompting
- Occasional prompting
- Few-shot prompting
- Zero-shot prompting
Explanation: When using few-shot prompting, the goal is to aid the large language model (LLM) in producing the required output by offering a limited number of instances or demonstrations inside the prompt itself. In this particular instance, the graphic designer provides three examples to highlight the context, language, and desired structure of the overview of the most recent trends in the design business. Through the use of these samples, the artist provides the LLM with some preliminary direction on the kind of information that is anticipated and how it need to be provided. As a result, the LLM can have a better understanding of the work, which in turn increases the likelihood that they will provide output that is correct and relevant. Some tasks, such as those in which the required output can be inferred from a minimal quantity of sample data, are good candidates for the approach known as few-shot prompting.
