Retrieval Augmented Generation (RAG)

Often these Answer Engines use a technique called RAG (Retrieval Augmented Generation) Sequence. One of the coolest uses of Large Language Models (LLMs) is creating smart question-answering (Q&A) chatbots. These are tools that can answer questions based on specific information. They use a method called Retrieval Augmented Generation, or RAG.

How Answer Engines Interact with Foundational Models?

Ever wondered what happens under the hood when you ask your question to an answer engine? Below is a great representation of how the web application layer interacts with the foundational model and retrieves the information. This uses something called the RAG Architecture (quoted in the second image below)