AI Wiki · Mediocre Security
Concept

Retrieval-augmented generation (RAG)

Having a model look up relevant documents at query time and answer from them, instead of relying only on what it memorized in training.

RAG retrieves relevant documents (usually via embeddings in a vector database) and puts them in the model’s context at query time. It grounds answers in current, private data and reduces hallucination. It also imports untrusted content into the prompt, making it a common prompt-injection path; hidden prompts in WhatsApp help pages show how content can be written for the machines that retrieve it. Tools: JamAI Base, Perplexity’s pplx-embed models, the Document Parsing Protocol.

Connections

Retrieval-augmented generation (RAG) and everything one link away. Drag to move, scroll to zoom, click a node to open it.Open in the full graph →

What links here

Concept Context window · Embedding · Hallucination · Vector database

Product JamAI Base

Company LlamaIndex

Protocol LlamaIndex Document Parsing Protocol

Sources

  1. Wikipedia, “Retrieval-augmented generation”