How many Refer to the Exhibit Practice Questions questions are on this page?

This page has 12 Refer to the Exhibit Practice Questions scenario questions for the Generative AI Leader exam, each with detailed explanations and wrong-answer analysis.

How should I approach Generative AI Leader scenario questions?

Read the full scenario before looking at the answer options. Identify the constraint or requirement in the scenario, then eliminate options that are generally true but wrong for this specific case. Scenario questions reward careful reading over pattern matching.

← Back to Google Cloud Generative AI Leader Generative AI Leader questions

Scenario-based practice

Refer to the Exhibit Practice Questions

Practise Google Cloud Generative AI Leader Generative AI Leader practice questions — original exam-style scenarios covering every exam domain, with detailed explanations, wrong-answer analysis, and common exam traps.

Start full practice test Read exam guide

scenario questions

Generative AI Leader

exam code

Google Cloud

vendor

Scenario guide

How to approach refer to the exhibit practice questions

Practise exhibit-style questions that ask you to read a topology, table, command output or diagram before choosing the best answer.

Quick answer

Exhibit-style questions test whether you can read a topology, command output, diagram or table before choosing the best answer.

How to extract the relevant detail from an exhibit.

How topology, command output or routing information affects the answer.

How to avoid answering from memory before reading the evidence.

How to map the exhibit back to the exam objective.

Practice scenarios

Question 1mediummultiple choice

Full question →

Refer to the exhibit. The endpoint is experiencing high latency during traffic spikes. The team wants to improve response time by reducing queueing. Which change to the configuration would be most effective?

Exhibit

Refer to the exhibit.

```
# Vertex AI Endpoint configuration
{
  "model": "gemini-1.5-pro",
  "endpoint": "projects/my-project/locations/us-central1/endpoints/123456789",
  "deployedModel": {
    "modelVersion": "1",
    "minReplicaCount": 1,
    "maxReplicaCount": 5,
    "autoscalingMetricSpecs": [
      {
        "metricName": "custom.googleapis.com|genai|request_count",
        "target": 100
      }
    ]
  }
}
```

A
Decrease minReplicaCount to 0
Why wrong: Fewer minimum replicas can increase cold starts.
B
Change the model version to '2'
Why wrong: Model version doesn't affect latency directly.
C
Decrease the target value in autoscaling metric to 50
Why wrong: Lower target triggers scaling earlier but may not reduce queueing if max replicas are insufficient.
D
Increase maxReplicaCount to 10
More replicas handle higher load.

Full breakdown with real-world context →

Question 2easymultiple choice

Full question →

The exhibit shows a command to deploy a model to a Vertex AI endpoint with GPU. The deployment fails due to a resource constraint. What is the most likely reason?

Exhibit

Refer to the exhibit.
```
Deploying a model to Vertex AI Endpoint with GPU:
$ gcloud ai endpoints deploy-model $ENDPOINT_ID \
  --model=$MODEL_ID \
  --machine-type=n1-standard-4 \
  --accelerator=count=1,type=nvidia-tesla-t4 \
  --min-replica-count=2 \
  --max-replica-count=5
```

A
The --model flag points to an autoML model.
Why wrong: AutoML models can be deployed to endpoints.
B
The accelerator type is misspelled.
Why wrong: nvidia-tesla-t4 is correct.
C
The machine type n1-standard-4 does not support GPU accelerators.
n1-standard machines do not have enough PCIe lanes; use n1-highmem or n1-highcpu.
D
The min-replica-count is greater than the max-replica-count.
Why wrong: min=2, max=5 is valid.

Full breakdown with real-world context →

Question 3hardmultiple choice

Full question →

The exhibit shows the deployment configuration for a conversational AI model used in a finance application. Users report that responses are creative but often contain factually incorrect financial advice. Which parameter change would most improve factual accuracy?

Exhibit

Refer to the exhibit.

```json
{
  "model": "publishers/google/models/chat-bison@001",
  "endpoint": "us-central1-aiplatform.googleapis.com",
  "parameters": {
    "temperature": 0.9,
    "topP": 0.95,
    "maxOutputTokens": 256,
    "groundingConfig": {
      "sources": []
    }
  },
  "deployment": "production"
}
```

A
Add grounding sources, such as "EnterpriseSearch" or "Web"
Grounding forces the model to base responses on real data, directly improving factual accuracy.
B
Lower temperature to 0.1
Why wrong: Lower temperature reduces randomness but does not add factual information; model may still hallucinate.
C
Increase topP to 1.0
Why wrong: Higher topP increases token sampling diversity, potentially worsening factual errors.
D
Increase maxOutputTokens to 1024
Why wrong: Changing token limit does not affect the factual correctness of content.

Full breakdown with real-world context →

Question 4easymultiple choice

Full question →

Refer to the exhibit. A company has this IAM policy on a Vertex AI project. Alice complains she cannot create a new model. What is the most likely reason?

Exhibit

Refer to the exhibit.
```
{
  "bindings": [
    {
      "role": "roles/aiplatform.user",
      "members": ["user:alice@example.com"]
    },
    {
      "role": "roles/aiplatform.admin",
      "members": ["user:bob@example.com"]
    }
  ]
}
```

A
She needs the roles/aiplatform.modelUser role
Why wrong: modelUser allows using but not creating models.
B
She needs the roles/aiplatform.admin role
Admin role has full access, including model creation.
C
She needs the roles/aiplatform.modelAdmin role
Why wrong: modelAdmin is not a built-in role; admin is the correct role.
D
She needs the roles/aiplatform.modelCreator role
Why wrong: modelCreator is not a built-in role; admin is needed.

Full breakdown with real-world context →

Question 5hardmultiple choice

Full question →

Refer to the exhibit. An administrator creates this IAM policy for a Vertex AI project. What is the effect of this policy?

Exhibit

{
  "bindings": [
    {
      "role": "roles/aiplatform.user",
      "members": ["user:alice@example.com"]
    },
    {
      "role": "roles/aiplatform.customCodeModelAdmin",
      "members": ["user:bob@example.com"]
    }
  ]
}

A
Alice can view models; Bob can delete models
Why wrong: aiplatform.user includes more than view; delete is not granted by customCodeModelAdmin.
B
Alice can deploy pre-trained models; Bob can create and manage custom model code
aiplatform.user includes deployment permissions; customCodeModelAdmin covers custom code management.
C
Both have full access to all Vertex AI resources
Why wrong: Neither role grants full access; aiplatform.user is limited, customCodeModelAdmin is specific.
D
Alice can train models; Bob can deploy models
Why wrong: Training is not typically part of aiplatform.user; customCodeModelAdmin includes deployment of custom models, but not necessarily all models.

Full breakdown with real-world context →

Question 6easymultiple choice

Full question →

Refer to the exhibit. A user wants formal translations from a generative AI model, but the model outputs informal style inconsistently. Which prompt engineering technique would best ensure consistent formal translations?

Exhibit

User: Translate the following sentence to French: Hello, how are you?
Model: Bonjour, comment vas-tu?
Expected: Bonjour, comment allez-vous? (formal)

A
Use context caching
Why wrong: Caching does not alter model behavior.
B
Provide a few-shot example with formal and informal pairs
Correct: Few-shot examples directly show the expected output format.
C
Use a longer system prompt with detailed rules
Why wrong: While helpful, detailed rules may not be as effective as examples for style.
D
Set top_k to 1
Why wrong: This makes output deterministic but does not enforce formal tone.

Full breakdown with real-world context →

Question 7mediummultiple choice

Full question →

Refer to the exhibit. A developer creates a model resource with this YAML config but gets an error that the model is not deployable. What is missing?

Exhibit

model:
  name: "projects/my-project/locations/us-central1/models/1234"
  explanation_spec:
    metadata:
      inputs:
        my_input:
          input_tensor_name: "input"
          modality: "text"

A
model_type
Why wrong: model_type is not a required field.
B
artifact_uri
Why wrong: artifact_uri is needed for custom models, but the primary requirement is container_spec for deployment.
C
container_spec
container_spec is required to tell Vertex AI which container to use.
D
description
Why wrong: Description is optional, not required for deployability.

Full breakdown with real-world context →

Question 8easymultiple choice

Full question →

Refer to the exhibit. What is the most likely cause of this error?

Exhibit

ERROR: (gcloud.aiplatform.models.upload) PERMISSION_DENIED: Permission 'aiplatform.models.upload' denied on resource project my-project

A
The user does not have the required IAM role
The error is a permission denial, meaning the user's IAM roles do not include 'aiplatform.models.upload'.
B
The model is too large
Why wrong: Size limits would cause a different error (e.g., quota exceeded).
C
The network is down
Why wrong: Network issues would typically result in timeout or connectivity errors, not permission denied.
D
The project ID is incorrect
Why wrong: An incorrect project ID would produce a different error (e.g., project not found).

Full breakdown with real-world context →

Question 9mediummultiple choice

Full question →

The exhibit shows the output of describing a model on Vertex AI. What does 'modelSource: MODEL_GARDEN' indicate about this model?

Exhibit

Refer to the exhibit.
```gcloud output
$ gcloud ai models describe --region=us-central1 my-model@123
Model:
  displayName: my-model
  versionId: "123"
  modelSource: MODEL_GARDEN
  supportedExportFormats:
  - id: "json"
  supportedInputStorageFormats:
  - json
  supportedOutputStorageFormats:
  - json
```

A
The model was imported from the Vertex AI Model Garden.
MODEL_GARDEN indicates it's a Model Garden model.
B
The model was trained on Vertex AI from scratch.
Why wrong: Model Garden models are pre-trained, not trained from scratch.
C
The model has been exported to Model Garden.
Why wrong: Export formats are for deployment, not Model Garden.
D
The model was fine-tuned using AutoML.
Why wrong: Fine-tuning would show different source, not MODEL_GARDEN.

Full breakdown with real-world context →

Question 10hardmultiple choice

Full question →

Refer to the exhibit. A data scientist is fine-tuning a model. The training loss and accuracy are improving each epoch. However, after training, the model performs poorly on a held-out validation set. What is the most likely issue?

Exhibit

Refer to the exhibit.
```
INFO: Training started.
INFO: Epoch 1/10 - loss: 2.3456 - accuracy: 0.1234
INFO: Epoch 2/10 - loss: 1.9876 - accuracy: 0.2345
INFO: Epoch 3/10 - loss: 1.6543 - accuracy: 0.3456
INFO: Epoch 4/10 - loss: 1.4321 - accuracy: 0.4567
INFO: Epoch 5/10 - loss: 1.2345 - accuracy: 0.5678
INFO: Epoch 6/10 - loss: 1.0987 - accuracy: 0.6789
INFO: Epoch 7/10 - loss: 0.9876 - accuracy: 0.7890
INFO: Epoch 8/10 - loss: 0.8765 - accuracy: 0.8901
INFO: Epoch 9/10 - loss: 0.7654 - accuracy: 0.9012
INFO: Epoch 10/10 - loss: 0.6543 - accuracy: 0.9123
```

A
Underfitting
Why wrong: Underfitting would show poor training performance as well.
B
Inappropriate learning rate
Why wrong: An inappropriate learning rate would typically cause instability or poor convergence, not this pattern.
C
Data leakage
Why wrong: Data leakage could cause overly optimistic validation scores, but here validation is poor.
D
Overfitting
Overfitting leads to good training performance but poor validation.

Full breakdown with real-world context →

Question 11easymultiple choice

Full question →

Refer to the exhibit. A user receives this error when trying to get predictions from a Vertex AI endpoint. What is the most likely cause?

Exhibit

ERROR: (gcloud.ai.platform.predict) PERMISSION_DENIED: Permission 'aiplatform.endpoints.predict' denied on resource 'projects/my-project/locations/us-central1/endpoints/123456' (or resource may not exist).

A
The endpoint does not exist
Why wrong: Would show NOT_FOUND, not PERMISSION_DENIED.
B
The endpoint is in a different region
Why wrong: Would produce a different error.
C
The user lacks necessary IAM permissions
PERMISSION_DENIED indicates missing permissions.
D
The model is not deployed
Why wrong: Would produce a different error.

Full breakdown with real-world context →

Question 12hardmultiple choice

Full question →

Refer to the exhibit. A developer receives this error when trying to call a model for prediction. What is the most likely cause?

Exhibit

{
  "error": {
    "code": 403,
    "message": "Permission 'aiplatform.models.predict' denied on resource 'projects/my-project/locations/us-central1/models/456'"
  }
}

A
The project has exceeded its prediction quota.
Why wrong: Quota exceeded usually returns a 429 or 503 status.
B
The developer's service account lacks the required IAM role.
The 403 error is a standard permission denied response from IAM.
C
The model version has been deprecated.
Why wrong: Deprecated models may still allow prediction but could return a warning; not a 403.
D
The model is not deployed on an endpoint.
Why wrong: That would typically result in a 404 or model not found error.

Full breakdown with real-world context →

These Generative AI Leader practice questions are part of Courseiva's free Google Cloud certification practice question bank. Courseiva provides original exam-style Generative AI Leader questions with detailed explanations, topic-based practice, mock exams, readiness tracking, and study analytics.

Refer to the Exhibit Practice Questions

How to approach refer to the exhibit practice questions

Quick answer

Related Generative AI Leader topic practice pages

Fundamentals of Generative AI practice questions

Business Strategies for Generative AI Solutions practice questions

Google Cloud's Generative AI Offerings practice questions

Techniques to Improve Generative AI Model Output practice questions

Generative AI Leader fundamentals practice questions

Generative AI Leader scenario practice questions

Generative AI Leader troubleshooting practice questions

Practice scenarios

Refer to the exhibit. The endpoint is experiencing high latency during traffic spikes. The team wants to improve response time by reducing queueing. Which change to the configuration would be most effective?

Exhibit

The exhibit shows a command to deploy a model to a Vertex AI endpoint with GPU. The deployment fails due to a resource constraint. What is the most likely reason?

Exhibit

The exhibit shows the deployment configuration for a conversational AI model used in a finance application. Users report that responses are creative but often contain factually incorrect financial advice. Which parameter change would most improve factual accuracy?

Exhibit

Refer to the exhibit. A company has this IAM policy on a Vertex AI project. Alice complains she cannot create a new model. What is the most likely reason?

Exhibit

Refer to the exhibit. An administrator creates this IAM policy for a Vertex AI project. What is the effect of this policy?

Exhibit

Refer to the exhibit. A user wants formal translations from a generative AI model, but the model outputs informal style inconsistently. Which prompt engineering technique would best ensure consistent formal translations?

Exhibit

Refer to the exhibit. A developer creates a model resource with this YAML config but gets an error that the model is not deployable. What is missing?

Exhibit

Refer to the exhibit. What is the most likely cause of this error?

Exhibit

The exhibit shows the output of describing a model on Vertex AI. What does 'modelSource: MODEL_GARDEN' indicate about this model?

Exhibit

Refer to the exhibit. A data scientist is fine-tuning a model. The training loss and accuracy are improving each epoch. However, after training, the model performs poorly on a held-out validation set. What is the most likely issue?

Exhibit

Refer to the exhibit. A user receives this error when trying to get predictions from a Vertex AI endpoint. What is the most likely cause?

Exhibit

Refer to the exhibit. A developer receives this error when trying to call a model for prediction. What is the most likely cause?

Exhibit