What is the main advantage of Retrieval-Augmented Generation (RAG)?

• A. It reduces model size.
• B. It improves factuality and allows up-to-date or domain-specific information.
• C. It eliminates the need for any prompts.
• D. It guarantees perfect citations.

Answer: (B)

Retrieval-Augmented Generation blends a neural generator with a retrieval component that pulls in relevant documents from an external corpus and conditions the answer on those documents. This grounding makes responses more factual and adaptable to information that is up-to-date or specific to a domain, since the model can rely on real sources rather than only its internal parameters. You can ask about recent events or specialized topics and get answers anchored to concrete materials. The other options aren’t the main benefit: reducing model size isn’t a consequence of retrieval, prompts are still needed, and while RAG can improve traceability, it doesn’t guarantee perfect citations because retrieved content can be misinterpreted or incomplete.