What Is Retrieval Augmented Generation (RAG)?

A technique that supplements a model's prompt with relevant retrieved external data at inference time to improve accuracy and grounding.

Retrieval Augmented Generation (RAG) across 2 exams

Retrieval Augmented Generation (RAG) appears on the following exams. Each defines it in the context candidates are tested on:

AWS AI Practitioner
A technique that supplements a model's prompt with relevant retrieved external data at inference time to improve accuracy and grounding.
AI-901
A technique that grounds generative model responses by retrieving relevant external data and including it in the prompt context.