« Back to Glossary Index

A technique used to improve the quality of output from AI models like OpenAI’s GPT-4. It involves a sequence of prompts or questions that progressively build on the model’s previous responses, refining the context and enhancing its understanding to produce more accurate and relevant results.