LLM Retrieval Augmented Generation
LLM Retrieval Augmented Generation — Grounding model answers in retrieved documents.
Definition
LLM Retrieval-Augmented Generation is RAG applied specifically to large language models: retrieve passages, stuff or rank them into context, then generate an answer that cites those passages.
Why it matters
It is the practical form of grounding used in Memory/Codex search. Quality depends on chunking, embedding recall, and whether the retrieved text actually contains a definition — not on model size alone.
Case
Wiki Ask RAG that only retrieves “Title — AI concept.” will paraphrase that stub. The failure is corpus thinness, not “the model is dumb.” Enrich the entry with Definition/Case, re-embed, then re-ask.
How to read it
Inspect top citations before trusting the prose. Prefer answers that start with a plain definition and admit missing evidence over long Alpha Factory narratives built on empty stubs.
Ask the macro AI about this object
Opens ZChat with Codex, RAG, and chart context enabled. Connected to the shared Ztrader memory layer.