Generative AI Leader Timed Practice Test 3 — 10 Questions

Question 1

A company wants to use Generative AI for customer support chatbots. They are concerned about cost and latency. Which deployment option best balances these concerns?

Accepted Answer

Use a fine-tuned version of a smaller model on Vertex AI with response caching. Option D is correct because using a fine-tuned smaller model on Vertex AI with response caching reduces both cost and latency. Smaller models require fewer computational resources, and caching avoids redundant inference calls, directly addressing the company's concerns without sacrificing accuracy for the specific task.

Answer

Deploy an open-source model on-premise to avoid cloud costs

Answer

Rely on a third-party chatbot API that abstracts the model

Answer

Use the largest available foundation model via API for highest accuracy

Question 2

Which TWO of the following are key differences between generative AI and discriminative AI? (Choose two.)

Accepted Answer

Generative models can create new data samples, while discriminative models only assign labels to existing data.. Option A is correct because generative AI models learn the underlying distribution of the data, enabling them to generate new, realistic samples (e.g., images, text) from the learned distribution. In contrast, discriminative models learn decision boundaries to classify or label existing data without the ability to create new data instances. This fundamental difference in capability—creation versus discrimination—is a core distinction between the two paradigms.

Answer

Generative models require less training data than discriminative models.

Answer

Generative models cannot be used for supervised learning tasks like classification.

Answer

Discriminative models always outperform generative models on tasks like image classification.

Question 3

A team uses Vertex AI to host a large language model. They want to reduce latency for real-time applications. What is the best strategy?

Accepted Answer

Use model quantization. Option C is correct because model quantization reduces the precision of the model's weights (e.g., from FP32 to INT8), which decreases memory footprint and computational requirements, directly lowering inference latency for real-time applications on Vertex AI. This is a standard optimization technique for deploying large language models with minimal accuracy loss while meeting latency SLAs.

Answer

Increase number of replicas

Answer

Switch to a smaller model

Answer

Use batch prediction instead of online

Question 4

Which THREE of the following are common techniques to reduce harmful biases in generative AI models? (Choose three.)

Accepted Answer

Use reinforcement learning from human feedback (RLHF) with a reward model that penalizes biased or unfair outputs.. A is correct because RLHF uses a reward model trained on human preferences to score model outputs, and explicitly penalizing biased or unfair outputs during fine-tuning directly reduces harmful biases. This technique aligns the model's behavior with human values by optimizing against a learned reward signal that captures bias-related concerns.

Answer

Decrease the model's temperature parameter to make outputs more deterministic.

Answer

Conduct a legal review of all generated outputs before release.

Question 5

A company is fine-tuning a Gemma model using Vertex AI. They observe that the model overfits. Which TWO actions should they take to mitigate overfitting?

Accepted Answer

Use more diverse data. Option C is correct because introducing more diverse data helps the model generalize better by exposing it to a wider variety of patterns, reducing the risk of memorizing noise from a limited dataset. Option E is correct because dropout randomly deactivates a fraction of neurons during fine-tuning, which prevents co-adaptation and acts as a regularization technique to combat overfitting in transformer-based models like Gemma.

Answer

Use a larger batch size

Answer

Increase the number of training epochs

Answer

Reduce the learning rate

Question 6

Which THREE of the following are potential risks when deploying generative AI?

Accepted Answer

Hallucinations. Option A is correct because generative AI models, particularly large language models (LLMs), can produce plausible-sounding but factually incorrect or nonsensical outputs, known as hallucinations. This occurs due to the model's probabilistic nature and lack of true understanding, where it generates text based on learned patterns rather than verified facts.

Answer

Memorization of sensitive training data

Answer

Increased model accuracy

Question 7

A company is using a generative AI model to generate product descriptions. They notice the outputs often include factual inaccuracies about product specifications. Which technique would best address this issue without modifying the model's architecture?

Accepted Answer

Implement a Retrieval-Augmented Generation (RAG) pipeline that retrieves product specs from a database. Retrieval-Augmented Generation (RAG) is the correct technique because it grounds the model's output in factual, up-to-date product specifications retrieved from an external database. This directly addresses factual inaccuracies without modifying the model's architecture, as the model generates text based on retrieved context rather than relying solely on its parametric knowledge.

Answer

Decrease the temperature parameter to 0.1

Answer

Increase the max output tokens to 1024

Answer

Use few-shot prompting with 5 examples of correct descriptions

Question 8

What is the primary purpose of a system instruction in the Gemini API?

Accepted Answer

Define the overall behavior and constraints for the model. The system instruction in the Gemini API is the primary mechanism to define the overall behavior, persona, constraints, and guardrails for the model across all interactions. Unlike per-query parameters, it sets a persistent context that shapes how the model interprets every user prompt, ensuring consistent adherence to rules such as tone, format, or safety policies.

Answer

Set the model's temperature and top_p

Answer

Provide few-shot examples for each query

Answer

Set the maximum output length

Question 9

Which TWO techniques are effective for reducing bias in generative AI model outputs?

Accepted Answer

Training on diverse and representative datasets. Option B is correct because training on diverse and representative datasets directly reduces sampling bias and coverage gaps in the training distribution, which are primary sources of stereotypical or skewed outputs. By ensuring the model sees balanced examples across demographics, contexts, and edge cases, it learns more equitable representations and reduces the likelihood of generating biased content.

Answer

Increasing model size to learn more patterns

Answer

Relying solely on post-hoc filters

Answer

Limiting the model to only factual prompts

Question 10

What are THREE benefits of using embedding models in a Retrieval Augmented Generation (RAG) system?

Accepted Answer

They compress text into dense vectors for efficient retrieval.. Option A is correct because embedding models convert text into dense vector representations that capture semantic meaning, enabling efficient similarity search in vector databases. This compression reduces the dimensionality of the data, allowing the RAG system to quickly retrieve the most relevant documents from a large corpus based on vector distance metrics like cosine similarity.

Answer

They allow the model to generate new training data automatically.

Answer

They provide deterministic outputs for the same query.