Retrieval-Augmented Generation (RAG).md

created	modified	tags	type	status
2024-07-15T10:13	2024-10-16 12:58	llm rag large-language-model	note	completed

Retrieval-Augmented Generation (RAG) refers to the process of including additional contextual information in the prompt of a generative language model, fetched from an external knowledge base (information source). This gives the language model direct access to information which might answer the user's query, rather than needed to access it's own "knowledge.", greatly helping to reduce hallucination.

Related

• [[Advanced RAG Techniques]]
• [[Applied Large Language Model Concepts]]
• [[Graph RAG]]
• [[Searching for Best Practices in Retrieval-Augmented Generation (paper)]]
• [[happy-rag-friends (github repo)]]