Generative AI Leader Practice Test 37 — 15 Questions

Question 1

A company uses Vertex AI Agent Builder to create a customer support agent. They need the agent to answer questions about order status by calling an internal API. Which Vertex AI feature should they use?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Vertex AI Extensions. Extensions in Vertex AI Agent Builder allow the agent to call external APIs (including internal ones) as tools during conversation.

Answer

Vertex AI RAG Engine

Answer

Grounding with Google Search

Answer

Vertex AI Model Garden

Question 2

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 3

A developer runs the command above to test a text classification model deployed on a Vertex AI endpoint. The model returns an error. What is the most likely cause?

Accepted Answer

The instance schema (e.g., 'content' field) does not match the model's expected input signature. The most common cause of inference errors on Vertex AI endpoints is a mismatch between the input instance schema (e.g., the 'content' field in the JSON request) and the model's expected input signature. Vertex AI validates the request payload against the model's saved signature (typically from TensorFlow SavedModel or PyTorch TorchScript), and if the field names, types, or shapes do not match, the endpoint returns an error rather than a prediction.

Answer

The endpoint ID '789' does not exist in the project

Answer

The model is not deployed to any endpoint

Answer

The region 'us-central1' does not match the region where the model is deployed

Question 4

An organization needs to deploy a generative AI application with strict compliance requirements, including data residency and auditability of model decisions. Which Google Cloud feature should they prioritize?

Accepted Answer

Vertex AI. Vertex AI provides enterprise controls including data residency options, audit logs, and model monitoring for compliance. Other options are important but do not directly address data residency and auditability comprehensively.

Answer

Colab Enterprise

Answer

Gemini API

Answer

Model Garden

Question 5

A developer wants to add real-time speech transcription to a customer call center application. They need low latency and high accuracy for multiple languages. Which Google AI API is most appropriate?

Accepted Answer

Speech-to-Text API. The Speech-to-Text API is the correct choice because it is specifically designed to convert audio into text in real time, supporting over 125 languages and variants with low-latency streaming. It offers features like automatic punctuation, speaker diarization, and domain-specific models (e.g., phone call) that directly meet the requirements of a customer call center application needing high accuracy across multiple languages.

Answer

Natural Language API

Answer

Text-to-Speech API

Answer

Translation API

Question 6

Which THREE steps are required to secure a generative AI pipeline that uses Vertex AI and involves sensitive customer data?

Accepted Answer

Use VPC Service Controls to create a perimeter around Vertex AI resources. VPC Service Controls are required to create a service perimeter around Vertex AI resources, preventing data exfiltration by restricting data movement across the perimeter boundary. This is critical for sensitive customer data because it mitigates the risk of unauthorized access or leakage, even from within the same project or organization.

Answer

Expose the prediction endpoint publicly with an API key

Answer

Disable audit logging to reduce data exposure

Question 7

A company is using a large language model to generate code reviews. They want to reduce token costs while maintaining quality. Which approach is MOST effective?

Accepted Answer

Use a smaller model fine-tuned for code (e.g., Codey). Using a smaller, specialized model for code tasks reduces cost per token and often performs well on code tasks. Caching is not applicable for code generation. Reducing max tokens might truncate output. Fewer examples may harm quality.

Answer

Reduce the max output tokens to 100

Answer

Use a zero-shot prompt instead of few-shot to reduce prompt tokens

Answer

Cache frequently used code snippets in a prompt template

Question 8

A company needs to deploy a generative AI application on Google Cloud that meets data residency requirements. Which THREE features should they enable? (Select three.)

Accepted Answer

Enable VPC Service Controls. To satisfy data residency, the company must control where data is stored (region selection), ensure no data leaves that region (data boundaries), and use a service that supports regional endpoints (Vertex AI).

Answer

Use a multi-region storage bucket

Answer

Use the global endpoint for low latency

Question 9

A startup wants to quickly prototype a gen AI application. Which Google Cloud service should they use first?

Accepted Answer

Gen AI Studio. Gen AI Studio (now part of Vertex AI) provides a low-code/no-code interface for quickly prototyping generative AI applications using pre-trained models like PaLM 2 and Gemini. It allows startups to experiment with prompts, tune models, and deploy without managing infrastructure, making it the fastest path from idea to prototype.

Answer

Vertex AI Workbench

Answer

Cloud TPUs

Answer

Dataflow

Question 10

A developer uses the Vertex AI Python SDK to call a Gemini model for structured JSON output. However, the model often returns malformed JSON. Which parameter should the developer set in the generation configuration to enforce valid JSON output?

Accepted Answer

Set the 'response_mime_type' parameter to 'application/json'.. Setting `response_mime_type` to `'application/json'` in the generation configuration instructs the Gemini API to constrain the model's output to valid JSON format. This parameter leverages the model's native structured output capability, ensuring the response adheres to JSON syntax without relying on post-processing or prompt engineering.

Answer

Set the temperature to a lower value (0.1) to reduce variation.

Answer

Include few-shot examples of the desired JSON format in the system prompt.

Answer

Switch to a smaller model to reduce complexity.

Question 11

A company is deploying a code generation assistant for internal developers. They want to ensure the generated code is secure and follows best practices. Which two Vertex AI features should they use? (Choose TWO)

Accepted Answer

Grounding with Google Search for real-time security best practices. Grounding with Google Search can retrieve up-to-date security best practices, and Model Evaluation can assess code quality. The other options are not directly relevant to security or best practices.

Answer

Vertex AI Agent Builder for conversation

Answer

Cloud DLP for data loss prevention

Answer

AutoML Image for code snippet images

Question 12

A machine learning team is using Vertex AI to train a custom model. They want to optimize hyperparameters automatically. Which TWO steps are necessary to set up hyperparameter tuning in Vertex AI? (Choose TWO)

Accepted Answer

Define a hyperparameter metric in the training code. To run hyperparameter tuning, you must specify a hyperparameter metric in the training code and configure the tuning job in Vertex AI with the parameter specifications.

Answer

Enable Vertex AI Experiments

Answer

Use a custom container with a GPU

Answer

Enable distributed training across multiple nodes

Question 13

A medical imaging team wants to generate synthetic X-ray images to augment a training dataset for a rare disease. Which type of generative model is most suitable for generating high-fidelity, realistic medical images?

Accepted Answer

Diffusion model. Diffusion models are the most suitable for generating high-fidelity, realistic medical images because they iteratively denoise random noise into a coherent image through a learned reverse diffusion process, which produces superior sample quality and diversity compared to GANs, especially for complex, high-dimensional data like X-rays. Their training stability and ability to model fine-grained anatomical details without mode collapse make them the current state-of-the-art for medical image synthesis.

Answer

Generative Adversarial Network (GAN)

Answer

Variational Autoencoder (VAE)

Answer

Autoregressive transformer (e.g., PixelCNN)

Question 14

Refer to the exhibit. A developer sees this error when trying to deploy a model from Vertex AI Model Registry. What is the most likely cause?

Accepted Answer

The developer used the model display name instead of the full resource name. The error occurs because Vertex AI Model Registry requires the full resource name (e.g., 'projects/{project}/locations/{region}/models/{model_id}') to deploy a model, not just the display name. The display name is a human-readable label that is not unique within a project, while the full resource name uniquely identifies the model version. Using the display name causes the API to fail with a 'not found' or 'invalid argument' error.

Answer

The region is not supported

Answer

The model is not published

Answer

The model is in a different project

Question 15

A healthcare startup is developing a generative AI system to assist doctors in diagnosing rare diseases. According to Google's AI Principles, what is the MOST important requirement before deployment?

Accepted Answer

The system must include a mechanism for human review of all diagnostic suggestions. Google's AI Principles state that AI systems should be built and tested for safety, especially in high-stakes domains like healthcare, where human oversight is critical.

Answer

The model must achieve at least 99% accuracy on a held-out test set

Answer

The model must be trained on the most recent medical literature

Answer

The startup must publish the model's architecture in a peer-reviewed journal