Technical
RAG (Retrieval-Augmented Generation)
An architecture that combines a retrieval system with an LLM — the model first retrieves relevant documents, then generates answers grounded in that retrieved context.
Explained at five levels
Level 1
When the AI looks up real information in a library before answering, instead of just guessing from memory.
Level 2
A technique where AI searches through actual documents or databases first, then uses what it finds to give you a more accurate answer.
Level 3
An architecture that combines a retrieval system with an LLM — the model first retrieves relevant documents, then generates answers grounded in that retrieved context.
Level 4
A hybrid approach pairing a retrieval module (dense or sparse) with a generative model, allowing the system to condition its output on dynamically retrieved evidence, reducing hallucination and enabling knowledge updates without retraining.
Level 5
A retrieve-then-generate pipeline where a bi-encoder or cross-encoder retriever supplies grounding documents to a conditional language model, decomposing world knowledge into a parametric component (model weights) and a non-parametric component (external index).
Definitions are educational summaries. Terminology can vary by source and context.
Sources
- NIST Trustworthy and Responsible AI Resource Center: Glossary — National Institute of Standards and Technology. Accessed 2026-07-20.