Glossary term
Glossary term
Agentic Systems
A prompt that contains more than one (a "few") example demonstrating how the large language model should respond. For example, the following lengthy prompt contains two examples showing a large language model how to answer a query.
Few-shot prompting generally produces more desirable results than zero-shot prompting and one-shot prompting. However, few-shot prompting requires a lengthier prompt.
Few-shot prompting is a form of few-shot learning applied to prompt-based learning.
See Prompt engineering in Machine Learning Crash Course for more information.
Created for this library
A SaaS team uses few-shot prompting with three labeled examples in the prompt to bootstrap a ticket classifier without fine-tuning.
A legal assistant tool uses few-shot prompting with two annotated clauses to coax the model toward the desired output format.
A research team uses few-shot prompting with carefully chosen examples to anchor the model on the right reasoning pattern for a new task.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License