Fundamentals of Generative AI →mediumMultiple Select

Generative AI Leader Fundamentals of Generative AI Practice Question

A healthcare startup is building a generative AI application to draft patient education materials. They need to ensure the outputs are accurate, up-to-date, and tailored to each patient's condition. Which two techniques should they use to ground the model's responses in reliable medical knowledge? (Choose two.)

⚠ Common exam trap

The trap here is assuming that fine-tuning or a larger context window alone can ground a model in reliable knowledge, when retrieval and validation are needed for accuracy and currency.

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

✓

Retrieval-augmented generation (RAG) using a curated medical knowledge base.

Retrieval-augmented generation (RAG) grounds the model by providing relevant, trusted documents as context, while a fact-checking layer validates outputs against a reliable API. Together, they ensure accuracy and currency. Fine-tuning, few-shot prompting, and larger context windows do not inherently guarantee factual correctness or access to the latest medical knowledge.

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

✓
Retrieval-augmented generation (RAG) using a curated medical knowledge base.
Why this is correct
RAG retrieves relevant documents from a trusted knowledge base and provides them as context to the model, ensuring responses are grounded in authoritative, current information. This reduces hallucinations and allows tailoring to specific conditions by retrieving patient-specific or condition-specific guidelines. It is a core technique for building accurate generative AI applications in specialized domains.
✗
Fine-tuning the model on a large dataset of historical patient education materials.
Why it's wrong here
Fine-tuning can adapt the model's style and terminology, but it does not guarantee factual accuracy or up-to-date information. Historical materials may contain outdated practices, and the model could still hallucinate. Fine-tuning alone is not a reliable grounding technique for ensuring correctness and currency in a fast-changing field like medicine.
✗
Using a model with a larger context window to include more patient history.
Why it's wrong here
A larger context window allows more input data, but it does not ensure the model uses that data accurately or retrieves the most relevant information. It can help include more patient history, but without a retrieval mechanism, the model may still ignore or misinterpret the context. It is not a grounding technique by itself.
✗
Prompt engineering with few-shot examples of desired outputs.
Why it's wrong here
Few-shot prompting can guide the model's format and tone, but it does not provide a mechanism to verify facts or access external knowledge. The model may still generate inaccurate statements. It is a useful technique for shaping outputs but not sufficient for grounding in reliable medical knowledge.
✓
Implementing a fact-checking layer that queries a trusted medical API for validation.
Why this is correct
A fact-checking layer that queries a trusted medical API can validate the model's outputs against authoritative sources, catching inaccuracies before they reach the user. This adds a verification step that ensures up-to-date and correct information. It complements RAG by providing an additional check, making it a robust grounding strategy.

Go deeper

Generative AI Leader Fundamentals of Generative AI Practice Question

Why each option matters

Option-by-option breakdown

Related to this question