Q42 — AWS AIF-C01 Ch.3

Question 42 of 100 | ← Chapter 3

A company wants to improve the consistency and quality of responses from its large language model (LLM) by granting the model access to external knowledge sources. Which technique satisfies this requirement with minimal development effort?

• A. Fine-tuning
• B. Retrieval-Augmented Generation (RAG) ✓
• C. In-context learning
• D. Prompt engineering

Correct Answer: B. Retrieval-Augmented Generation (RAG)

Explanation

To improve the consistency and quality of large language model (LLM) responses by leveraging external knowledge sources, consider each technique: A. Fine-tuning typically requires substantial data and compute resources and is optimized for specific tasks, making it suboptimal for minimizing development effort. B. Retrieval-Augmented Generation (RAG) enhances model generation capabilities using external knowledge bases, allowing the model to reference external knowledge during text generation—improving consistency and quality while being relatively easy to integrate into existing models with minimal development effort. C. In-context learning primarily improves understanding of current dialogue context rather than directly providing external knowledge. D. Prompt engineering optimizes model performance by adjusting input prompts but does not inherently enable direct access to or utilization of external knowledge sources. Therefore, Retrieval-Augmented Generation (RAG) satisfies the requirement with minimal development effort. The correct answer is B.