Generative AI Leader Google Cloud's Generative AI Offerings Practice Test 2 — 15 Questions

Question 1

An organization is using Vertex AI Agent Builder to create a customer service agent. They want the agent to be able to hand off to a human agent when it cannot answer a question. What should they configure in the agent's design?

Accepted Answer

Add an 'Escalation' intent that triggers a human handoff. Agent Builder supports 'Escalation' intent to hand off to human agents. Option B is wrong because fallback intent is for unrecognized input but not necessarily human handoff. Option C is wrong because confirmation is for confirming actions. Option D is wrong because slot filling is for collecting parameters.

Answer

Configure 'Slot filling' to collect more info

Answer

Implement a 'Confirmation' prompt for the user

Answer

Use a 'Fallback' intent to route to a human

Question 2

A company is using Vertex AI for multimodal generative AI to analyze images and text. They need to ensure that the model's outputs are auditable and can be traced back to the input data. Which feature should they enable?

Accepted Answer

Vertex AI Model Monitoring with Explainable AI. Option D is correct because Vertex AI Model Monitoring with Explainable AI provides feature attributions that map model predictions back to specific input features (e.g., pixels in images or tokens in text). This creates an auditable trail by quantifying how each input contributed to the output, enabling traceability for compliance and debugging. The other options lack the direct input-to-output attribution required for auditability.

Answer

Vertex AI Feature Store

Answer

Vertex AI Experiments

Answer

Cloud Logging

Question 3

A data scientist wants to fine-tune a foundation model from Vertex AI Model Garden on their custom dataset. They want to choose a cost-effective method that updates only a small subset of parameters. Which fine-tuning approach should they use?

Accepted Answer

Parameter-Efficient Fine-Tuning (PEFT) like LoRA. Option C is correct because Parameter-Efficient Fine-Tuning (PEFT) methods like LoRA (Low-Rank Adaptation) are specifically designed to update only a small subset of parameters (e.g., low-rank matrices injected into transformer layers) while keeping the majority of the foundation model frozen. This drastically reduces memory and compute costs compared to full fine-tuning, making it the most cost-effective choice for customizing a model from Vertex AI Model Garden on a custom dataset.

Answer

Full fine-tuning

Answer

Prompt tuning

Answer

Reinforcement Learning from Human Feedback (RLHF)

Question 4

Which TWO features are available in Vertex AI Agent Builder to enhance the conversational abilities of an agent? (Choose TWO.)

Accepted Answer

Slot filling. Slot filling is correct because it allows the agent to collect required parameters (slots) from the user during a conversation, enabling multi-turn interactions to fulfill complex requests. In Vertex AI Agent Builder, slot filling is a core feature for conversational agents, as it systematically prompts for missing information (e.g., date, location) until all necessary slots are filled, enhancing the agent's ability to handle dynamic user inputs.

Answer

Sentiment analysis

Answer

Code execution

Answer

Knowledge base integration

Question 5

Which THREE considerations are critical when deploying a generative AI model using Vertex AI Endpoints for a latency-sensitive application? (Choose THREE.)

Accepted Answer

Model size and architecture. Model size and architecture directly impact inference latency because larger models with more parameters require more computation per request. For latency-sensitive applications, choosing a smaller or distilled model (e.g., Gemma 2B vs. 27B) or using quantization can reduce response times. Vertex AI Endpoints serve the model as-is, so the model's inherent computational cost is the primary driver of per-request latency.

Answer

Number of model versions

Answer

Number of model instances

Question 6

A developer receives the above JSON response from a Vertex AI language model. The output content is correct, but the developer expected the model to not answer geography questions. What should the developer do to prevent the model from responding to geography queries?

Accepted Answer

Configure a safety filter for the 'Geography' category. Option C is correct because Vertex AI provides safety filters that can be configured to block model responses in specific categories, including a 'Geography' category. By adjusting the safety filter thresholds for this category, the developer can prevent the model from answering geography queries while still allowing it to respond to other topics. This is a direct and effective method to enforce content restrictions without modifying the model's underlying behavior.

Answer

Adjust the safety filter thresholds for the 'Toxic' category

Answer

Enable Vertex AI Grounding with a geography knowledge base

Answer

Add a system instruction to not answer geography questions

Question 7

A team deployed a custom generative AI model using KServe on Google Kubernetes Engine (GKE) with the above configuration. They notice that the model is taking longer than expected to respond. What is the most likely cause?

Accepted Answer

The model requires more than 1 GPU for acceptable performance. The configuration specifies 1 GPU, but the model requires more than 1 GPU for acceptable performance. KServe on GKE allocates GPU resources based on the `limits` field; if the model's inference workload exceeds the memory bandwidth or compute capacity of a single GPU, latency increases due to queuing and serialization. This is the most likely cause of the slow response time, as GPU-bound models are sensitive to under-provisioning.

Answer

The CPU resource limits are too low

Answer

The model is crashing due to insufficient memory

Answer

The container image is too large and takes time to pull

Question 8

You are a machine learning engineer at a healthcare startup. Your team has developed a generative AI model that summarizes patient medical records. The model is deployed on Vertex AI Endpoints using a custom container. You have configured the endpoint with a single n1-standard-4 machine (4 vCPUs, 15 GB memory) without accelerators. The model uses a small transformer architecture. During load testing with 50 concurrent requests, you observe that the average latency is 8 seconds, which exceeds the requirement of 2 seconds. Additionally, some requests time out after 10 seconds. You suspect the CPU is the bottleneck. You also notice that the model inference code uses TensorFlow but is not optimized for inference. Which action should you take to reduce latency?

Accepted Answer

Deploy the model on a machine with a GPU and use TensorRT for inference optimization.. Option D is correct because the CPU is identified as the bottleneck, and deploying on a GPU with TensorRT optimization directly addresses this by accelerating the TensorFlow inference. TensorRT optimizes the model graph and fuses layers, significantly reducing latency for transformer-based models, which is essential to meet the 2-second requirement.

Answer

Reduce the model size by pruning and quantization, then redeploy.

Answer

Increase the request timeout to 30 seconds to accommodate the latency.

Answer

Enable autoscaling to handle the load with multiple replicas.

Question 9

You are a generative AI architect for a large e-commerce company. Your team has built a product description generator using Vertex AI's text-bison model. The model is accessed via the Vertex AI API from a web application. You have set the temperature to 0.5 and top_k to 40. The team reports that the generated descriptions are often too generic and lack creativity. They want the descriptions to be more diverse and engaging. You are also concerned about cost, as each API call is billed. Which change should you recommend to increase creativity while managing cost?

Accepted Answer

Increase the temperature to 0.8 and keep top_k at 40.. Increasing the temperature to 0.8 makes the model's output probability distribution flatter, which increases randomness and allows less likely tokens to be selected. This directly addresses the need for more diverse and creative descriptions. Keeping top_k at 40 ensures the model still considers a broad set of candidate tokens, balancing creativity with coherence, and does not increase API call costs since temperature and top_k are inference parameters that do not affect billing.

Answer

Keep temperature at 0.5 but reduce top_k to 20.

Answer

Switch to a larger model like text-bison@002 and keep same parameters.

Answer

Decrease the temperature to 0.2 and increase top_k to 60.

Question 10

A global e-commerce company is using Vertex AI to build a generative AI chatbot for customer support. The chatbot is powered by the Gemini 1.5 Pro model and uses a vector search index for retrieval-augmented generation (RAG) over product documentation. The company has deployed the application in four regions (us-central1, europe-west4, asia-east1, and australia-southeast1) using a multi-region deployment with a global endpoint. The application is critical and requires high availability with a target latency of under 500ms for the RAG pipeline. Recently, users in Australia are experiencing inconsistent latency spikes, with response times exceeding 2 seconds during peak hours. The team suspects that the issue is related to the vector search index's replication and serving configuration. The index has 10 million embeddings with a dimension of 768. It is stored in a single regional bucket in us-central1, and the vector search index endpoint is deployed in all four regions with the same deployed index ID. The team is using the default configuration for index updates and serving. Which action should the team take to resolve the latency issue for Australian users?

Accepted Answer

Move the vector search index to a multi-regional Cloud Storage bucket (e.g., 'us') to reduce latency for index updates.. The latency spike for Australian users is caused by the vector search index being stored in a single regional bucket in us-central1. When the index is updated, the new embeddings must be rebuilt and streamed from us-central1 to the australia-southeast1 endpoint, introducing significant cross-region latency. Moving the index to a multi-regional Cloud Storage bucket (e.g., 'us') allows the index to be served from a location closer to all regions, reducing the update propagation time and improving consistency for Australian users.

Answer

Create a new regional bucket in australia-southeast1 and store a copy of the index there, then redeploy the vector search index endpoint to use the local bucket.

Answer

Deploy a separate vector search index endpoint for each region with its own index copy stored in a regional bucket in that region.

Answer

Increase the number of replicas for the vector search index in all regions to improve throughput and reduce latency.

Question 11

A company is building a customer support chatbot using Vertex AI Agent Builder. They want the agent to answer questions based on their internal knowledge base. Which feature should they use?

Accepted Answer

Grounding with enterprise data stores. Option B is correct because Vertex AI Agent Builder supports grounding with enterprise data stores, allowing the agent to answer from internal knowledge bases. Option A is wrong because Google Search grounding is for public data. Option C is wrong because model tuning adapts the model, not the data source. Option D is wrong because prompt engineering is used to shape responses, not to provide data.

Answer

Grounding with Google Search

Answer

Model tuning

Answer

Prompt engineering

Question 12

A data scientist is using Vertex AI Model-as-a-Service (MaaS) to deploy a fine-tuned open-source model. They notice high latency during inference. What is the most likely cause?

Accepted Answer

The endpoint is set to autoscaling with a low minimum node count. Option C is correct because a low minimum node count in autoscaling can cause cold starts and high latency. Option A is wrong because model size is managed by MaaS and typically handled. Option B is wrong because quantization affects model size and speed, but the issue is more likely autoscaling. Option D is wrong because region does not directly cause latency spikes.

Answer

The model is too large for the hardware

Answer

The model is not quantized

Answer

The region is incorrect

Question 13

A company is using Gemini Pro for code generation. They want to ensure that the generated code does not contain security vulnerabilities. Which approach should they implement?

Accepted Answer

Implement a human-in-the-loop review with automated scanning. Option D is correct because for security, human review combined with automated scanning is best practice. Option A is wrong because custom safety filters are generic and not code-specific. Option B is wrong because grounding with scanning tools is not a standard feature. Option C is wrong because safety settings are about content harm, not code vulnerabilities.

Answer

Enable grounding with security scanning tools

Answer

Use the Vertex AI Codey API with safety settings

Answer

Use a custom safety attribute filter

Question 14

You want to use a Google foundation model to generate text summaries of news articles. Which Vertex AI service should you use?

Accepted Answer

Vertex AI Generative AI Studio. Option B is correct because Vertex AI Generative AI Studio provides access to foundation models for text generation. Option A is wrong because Vertex AI Prediction is for deploying custom models. Option C is wrong because Model Registry manages model versions. Option D is wrong because Feature Store is for ML features.

Answer

Vertex AI Prediction

Answer

Vertex AI Model Registry

Answer

Vertex AI Feature Store

Question 15

A developer is using the Vertex AI Gemini API to generate product descriptions. They get a 400 error 'INVALID_ARGUMENT: The model's maximum input token limit is 8192.' What is the most likely issue?

Accepted Answer

The prompt is too long. Option A is correct because the error indicates the input prompt exceeds the token limit. Option B is wrong because the error says 'input token limit', not output. Option C is wrong because an invalid API key would give a different error (e.g., PERMISSION_DENIED). Option D is wrong because model availability would give a different error.

Answer

The API key is invalid

Answer

The output tokens are too high

Answer

The model is not available in the region