Courseiva

CCNA Fundamentals of Generative AI Questions

6 of 156 questions · Page 3/3 · Fundamentals of Generative AI · Answers revealed

151
MCQmedium

A marketing team at a retail company is using a generative AI model on Vertex AI to produce product descriptions from short bullet lists. They observe that the model's outputs are fluent but frequently invent specifications, such as claiming a jacket is waterproof when no such attribute was provided. The team wants to reduce these fabrications without retraining the model. Which approach should they take?

A.Lower the model's top-p value to 0.1 to force deterministic outputs.
B.Increase the model's temperature parameter to encourage more diverse outputs.
C.Increase the maximum output tokens to give the model more room to explain its reasoning.
D.Ground the model's responses by providing the bullet points as context in the prompt and instructing it to use only that information.
AnswerD

Grounding the model with the supplied bullet points as explicit context, combined with an instruction to rely solely on that information, constrains generation to the provided facts. This directly reduces fabricated attributes without retraining. It is the standard prompt-level technique for improving factual consistency in Vertex AI generative AI applications when the source data is available at inference time.

Why this answer

Grounding the generation in the provided bullet points by including them as context and instructing the model to use only that information is the most direct way to prevent invented specifications. It leverages the source data already available at inference time and requires no retraining. Parameters such as temperature, top-p, and output length affect style and randomness, not factual fidelity to a given source.

Exam trap

The trap here is assuming that tuning decoding parameters like temperature or top-p will eliminate hallucinations, when those settings only change randomness and cannot supply missing factual grounding.

152
Multi-Selectmedium

A company is designing a prompt engineering strategy for a customer service chatbot using Gemini. Which two practices are recommended for improving response quality? (Choose TWO)

Select 2 answers
A.Use chain-of-thought prompting
B.Always provide multiple examples in the prompt
C.Avoid any context in the prompt
D.Set temperature to 1.0 for maximum creativity
E.Include a system instruction to define the role
AnswersA, E

Chain-of-thought prompting improves response quality by having Gemini generate intermediate reasoning steps before the final answer, which raises accuracy on multi-step customer service queries such as troubleshooting or policy interpretation. This directly satisfies the stem's goal of improving response quality within the prompt engineering strategy.

Why this answer

Option A (Use chain-of-thought prompting) is correct because prompting Gemini to reason step by step before producing a final answer improves accuracy and coherence on complex customer service queries, especially those requiring multi-step logic or policy lookups. Option E (Include a system instruction to define the role) is correct because a system instruction sets persistent behavioral context, such as "You are a helpful customer service agent for X company," which anchors tone, scope, and constraints across all turns without repeating them in every user prompt. Option B is not recommended as a blanket rule: few-shot examples can help, but "always" providing multiple examples wastes tokens, increases latency, and can bias or overfit responses when zero-shot or single-example prompting suffices.

Option C is wrong because omitting context strips the model of the grounding information needed to answer domain-specific questions accurately. Option D is wrong because temperature 1.0 increases randomness and creativity, which is undesirable for a customer service chatbot where factual, consistent, and deterministic answers are required; lower temperatures (e.g., 0.2-0.4) are typically preferred.

Exam trap

Google Cloud often tests the misconception that higher temperature always improves creativity, but in customer service, lower temperature is critical for deterministic, safe responses, and candidates may overlook the role of system instructions in defining behavior.

153
MCQmedium

A media company is using a foundation model in Vertex AI to summarize long articles. They notice that summaries sometimes omit key details from the middle of very long articles. Which action should they take to improve summary completeness while staying within the model's context window?

A.Add a system instruction telling the model to always include all key details.
B.Split the article into chunks, summarize each chunk, then combine the summaries in a final prompt.
C.Switch to a smaller model variant to reduce latency and cost.
D.Increase the model's temperature setting to encourage more creative output.
AnswerB

Chunking and hierarchical summarization keeps each request within the context window and ensures middle content is processed. By summarizing segments separately and then synthesizing, the model can capture details that would otherwise be dropped due to long-context limitations. This approach is a standard pattern for handling documents that exceed a single prompt's capacity while maintaining completeness.

Why this answer

Long articles can exceed a model's effective context, causing middle content to be overlooked. Chunking the article, summarizing each part, and then combining those summaries ensures all sections are processed and represented. This hierarchical approach stays within context limits and improves completeness, directly addressing the observed omission of key details.

Exam trap

The trap here is thinking that prompt wording or temperature can force a model to retain content that falls outside its effective context window.

154
Multi-Selecthard

Which THREE are valid methods to reduce bias in generative AI outputs?

Select 3 answers
A.Using only English prompts
B.Increasing model size
C.Using a more diverse training dataset
D.Using safety filters
E.Applying prompt engineering to instruct the model to be fair
AnswersC, D, E

A more diverse training dataset directly addresses representation bias by exposing the model to broader demographic, cultural and linguistic variation during pre-training, reducing the likelihood of skewed associations in generated outputs. This satisfies the stem's requirement for a valid bias-reduction method, since bias frequently originates from unrepresentative or historically skewed training corpora.

Why this answer

Option C is correct because a more diverse training dataset reduces representational bias by exposing the model to a wider range of demographics, cultures, languages, and viewpoints, so the learned distribution is less skewed toward a dominant group. Option D is correct because safety filters (e.g., content moderation classifiers, toxicity detectors, or bias guardrails applied to prompts and outputs) can detect and block biased or harmful generations, mitigating bias at inference time. Option E is correct because prompt engineering that explicitly instructs the model to be fair, neutral, or inclusive (e.g., system prompts specifying balanced representation) steers generation toward less biased outputs without retraining.

Option A is not correct because restricting prompts to English narrows the linguistic and cultural scope, which can amplify bias rather than reduce it. Option B is not correct because increasing model size alone does not remove bias and can even amplify biases present in the training data.

Exam trap

Google Cloud often tests the misconception that increasing model size or using a single language (like English) can solve bias, when in reality these actions can worsen bias by amplifying existing skews or introducing new cultural blind spots.

155
MCQhard

A company is deploying a generative AI application that generates medical reports. They need to ensure the output is factual and minimizes hallucinations. Which approach is most effective?

A.Fine-tune the model with RLHF
B.Set the temperature to 0.0
C.Implement retrieval-augmented generation (RAG) with a curated knowledge base
D.Use prompt engineering to instruct the model to be accurate
AnswerC

RAG grounds generation in retrieved documents from a curated knowledge base, so the model conditions output on verified source content rather than parametric memory alone. This directly reduces hallucination and improves factual accuracy, satisfying the medical report requirement.

Why this answer

Retrieval-Augmented Generation (RAG) is the most effective approach because it grounds the model's output in a curated, authoritative knowledge base of medical data. By retrieving relevant, verified documents at inference time, RAG directly reduces the model's reliance on its parametric memory, which is the primary source of hallucinations in generative AI. This is especially critical in high-stakes domains like medical reporting, where factual accuracy is paramount.

Exam trap

The trap here is that candidates often choose 'Set the temperature to 0.0' because they confuse reducing randomness with eliminating factual errors, but temperature only controls output variability, not the truthfulness of the model's internal knowledge.

How to eliminate wrong answers

Option A is wrong because RLHF (Reinforcement Learning from Human Feedback) optimizes the model for human preference alignment and helpfulness, but it does not provide a mechanism to retrieve or verify facts from an external source, so it cannot reliably prevent hallucinations in factual domains. Option B is wrong because setting temperature to 0.0 makes the model deterministic (always picking the highest-probability token), but it does not correct factual errors stored in the model's weights; the model can still confidently generate false information. Option D is wrong because prompt engineering instructs the model to be accurate, but it is a soft constraint that the model can easily override; without external grounding, the model has no way to verify its own output against a trusted source.

156
MCQhard

A financial services company is using a large language model on Vertex AI to generate summaries of earnings reports. They notice that the model sometimes includes information not present in the source reports, leading to compliance risks. They want to reduce hallucinations by ensuring the model's output is grounded in the provided documents. Which technique should they implement?

A.Using a smaller model
B.Increasing the temperature
C.Retrieval-augmented generation (RAG)
D.Fine-tuning the model on earnings reports
AnswerC

RAG involves retrieving relevant passages from the source documents and providing them as context to the model, which reduces hallucinations by grounding the generation in factual content. This directly addresses the compliance risk by ensuring the summary is based on the earnings reports, not the model's internal knowledge.

Why this answer

Retrieval-augmented generation (RAG) is the most effective technique to ground model outputs in specific documents. By retrieving relevant passages from the earnings reports and including them in the prompt, the model is constrained to generate summaries based on that evidence, significantly reducing hallucinations and compliance risks.

Exam trap

The trap here is assuming fine-tuning eliminates hallucinations; it can adapt style but does not guarantee grounding in a specific input document.

← PreviousPage 3 of 3 · 156 questions total

Ready to test yourself?

Try a timed practice session using only Fundamentals of Generative AI questions.