Generative AI Leader Fundamentals of Generative AI Practice Question
A retail company uses the Vertex AI Gemini API to generate product descriptions. Recently, the model started producing factually incorrect statements about product specifications, such as wrong dimensions and materials. Which strategy should be implemented to improve factual accuracy?
⚠ Common exam trap
Candidates often mistakenly believe that fine-tuning the model or adjusting generation parameters (like temperature) can fix factual inaccuracies. However, these methods do not introduce new, verified data. The correct approach is grounding with Vertex AI Search, which uses retrieval-augmented generation (RAG) to pull authoritative product specifications from a trusted data source, directly addressing hallucination of facts like dimensions and materials.
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
✓
Use grounding with Vertex AI Search to retrieve verified product data
Grounding with Vertex AI Search connects the Gemini API to a verified, structured data source (e.g., product catalog), enabling the model to retrieve and cite factual specifications rather than relying solely on its training data. This directly addresses the hallucination of wrong dimensions and materials by providing a retrieval-augmented generation (RAG) mechanism that overrides the model's parametric knowledge with authoritative, up-to-date information.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Enable model versioning to automatically roll back to a previous version
Why it's wrong here
Rolling back to a previous model version restores earlier weights, but the inaccuracies reflect missing grounding rather than a regression, so the same errors persist. It is tempting because versioning suits deployments where a recent change degraded quality, and rollback would be correct in that specific regression scenario.
- ✗
Fine-tune the model on a dataset of product images and descriptions
Why it's wrong here
Fine-tuning on images and descriptions teaches style and format, not retrieval of verified specifications, so dimensions and materials remain ungrounded. It is tempting because fine-tuning adapts model behaviour, and it would be correct when the goal is consistent tone or domain-specific phrasing rather than factual grounding.
- ✗
Increase the temperature parameter to 0.9
Why it's wrong here
Raising temperature to 0.9 increases sampling randomness, which amplifies hallucination of specifications instead of constraining output to verified facts. It is tempting because temperature tuning is a recognised control, and a higher value would be correct when generating varied creative copy rather than accurate product data.
- ✓
Use grounding with Vertex AI Search to retrieve verified product data
Why this is correct
Grounding with Vertex AI Search anchors generation to retrieved, verified product data, so specifications come from an authoritative source rather than the model's parametric memory. This directly reduces the fabricated dimensions and materials the stem describes.
Go deeper
Related to this question
About these practice questions
One of 1,008 original Generative AI Leader practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. 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.