News for hackers. Stuff that matters.
Re:OpenAI GPT-4 vs. Groq Mistral-8x7B by samus on Friday March 22, 2024 @10:20AM (#39791011)
This is called Retrieval Augmented Generation (RAG). The LLM driver recognizes a query, it gets send to a vector database or to an external system (could be another LLM...) and the answer is placed in the context. It's a common strategy to work around their limited context length, but it tends to be brittle. Look for survey papers.

There are no comments on this comment yet.