Generative AI Leader Practice Question: Techniques to Improve Generative AI Model Output
To improve factuality in generative AI, which is the best approach?
⚠ Common exam trap
Google exams often test the misconception that hyperparameter tuning (like top_p) or scaling model size alone can fix factuality, when in reality these methods do not address the root cause of hallucination—lack of external grounding—and candidates may overlook the importance of retrieval and citation mechanisms, such as Google's Vertex AI Grounding or RAG.
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
✓
Grounded generation with citations
Grounded generation with citations directly addresses factuality by forcing the model to retrieve and cite verifiable sources (e.g., from a knowledge base or document store) before generating an answer. This approach, often implemented via retrieval-augmented generation (RAG), ensures outputs are anchored to external evidence rather than relying solely on the model's parametric memory, which can produce hallucinations. Citations also enable users to verify claims, making this the most effective technique for improving factual accuracy.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Set top_p to 0.1
Why it's wrong here
Low top_p narrows sampling to high-probability tokens, which increases consistency and reduces creative drift but does not supply external evidence, so the model can still assert confident falsehoods. It is the right lever when the scenario demands deterministic, repeatable phrasing rather than improved factuality.
- ✗
Reduce output length
Why it's wrong here
Shorter outputs do not add grounding, retrieval or verification, so unsupported claims persist at any length. It is tempting because brevity reduces the surface area for invented detail, and it genuinely helps when the goal is concise summarisation or latency reduction rather than factual accuracy.
- ✓
Grounded generation with citations
Why this is correct
Grounded generation retrieves authoritative source content and conditions the model's output on it, with citations letting users verify each claim. This directly constrains hallucination, satisfying the stem's goal of improving factuality rather than relying on the model's parametric memory alone.
- ✗
Increase model size
Why it's wrong here
Larger models memorise more and generalise better, yet scaling alone cannot guarantee currency or verifiability, and hallucination persists on niche or post-training facts. It is tempting because bigger models score higher on many benchmarks, and it is the correct choice when the constraint is capability rather than grounding.
Go deeper
Related to this question
About these practice questions
This Generative AI Leader question is part of Courseiva's 1,008-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This Generative AI Leader practice question is part of Courseiva's free Google Cloud certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the Generative AI Leader exam.