QUESTION 81
An AI practitioner is creating prompts for an Amazon Nova Canvas model. The AI practitioner wants to exclude certain characteristics from the model’s generated output. Which prompting technique will meet this requirement?

A. Mask prompts
B. Negative prompts
C. Inpainting prompts
D. Outpainting prompts

Correct Answer: B

QUESTION 82
Which statement correctly describes embeddings in generative AI?

A. Embeddings represent data as high-dimensional vectors that capture semantic relationships.
B. Embeddings is a technique that searches data to find the most helpful information to answer natural language questions.
C. Embeddings reduce the hardware requirements of a model by using a less precise data type for the weights and activations
D. Embeddings provide the ability to store and retrieve data for generative Al applications.

Correct Answer: A

QUESTION 83
A company has a generative Al application that uses a pre-trained foundation model (FM) on Amazon Bedrock. The company wants the FM to include more context by using company information. Which solution meets these requirements MOST cost-effectively?

A. Use Amazon Bedrock Knowledge Bases.
B. Choose a different FM on Amazon Bedrock.
C. Use Amazon Bedrock Agents.
D. Deploy a custom model on Amazon Bedrock.

Correct Answer: A

QUESTION 84
A company is using a pre-trained large language model (LLM) to build a chatbot for product recommendations. The company needs the LLM outputs to be short and written in a specific language. Which solution will align the LLM response quality with the company’s expectations?

A. Adjust the prompt
B. Choose an LLM of a different size.
C. Increase the temperature.
D. Increase the Top K value.

Correct Answer: A

QUESTION 85
A company wants to implement a single environment for both data and Al development Developers across different teams must be able to access the environment and work together. The developers must be able to build and share models and generative Al applications securely in the environment. Which AWS solution will meet these requirements?

A. Amazon Lex
B. Amazon SageMaker Unified Studio
C. Amazon Bedrock PartyRock
D. Amazon Q Developer

Correct Answer: B

QUESTION 86
A company is deploying a new AI application to generate content for internal users. Which strategy will make the application output more deterministic?

A. Decreasing the temperature
B. Increasing the learning rate
C. Setting stop sequences
D. Setting the token count

Correct Answer: A

QUESTION 87
What is the purpose of chunking in Retrieval Augmented Generation (RAG)?

A. To avoid database storage limitations for large text documents by storing parts or chunks of the text
B. To improve efficiency by avoiding the need to convert large text into vector embeddings
C. To improve the contextual relevancy of results retrieved from the vector index
D. To decrease the cost of storage by storing parts or chunks of the text

Correct Answer: C

QUESTION 88
A company wants to use Amazon SageMaker Al to help perform monthly customer satisfaction assessments. The task is not time sensitive. The company does not want to have persistent endpoints. Which inference option will meet these requirements MOST cost-effectively?

A. Real-time inference
B. Serverless inference
C. Batch transform
D. Asynchronous inference

Correct Answer: C

QUESTION 89
Which metric is used to evaluate the performance of foundation models (FMs) for text summarization tasks?

A. F1 score
B. Bilingual Evaluation Understudy (BLEU) score
C. Accuracy
D. Mean squared error (MSE)

Correct Answer: B

QUESTION 90
A company is preparing data to fine-tune a large language model (LLM). The company must follow the principles of responsible AI. Which actions will meet this requirement? (Select TWO.)

A. Include demographically diverse data sources to improve model fairness.
B. Apply consistent labeling standards across the training dataset.
C. Focus exclusively on high-volume data regardless of quality.
D. Use only synthetic data to train the model.
E. Use fewer labels to reduce ambiguity in data during model training.

Correct Answer: AB

QUESTION 91
A company is developing an ML model to make loan approvals. The company must implement a solution to detect bias in the model. The company must also be able to explain the model’s predictions. Which solution will meet these requirements?

A. Amazon SageMaker Clarify
B. Amazon SageMaker Data Wrangler
C. Amazon SageMaker Model Cards
D. AWS AI Service Cards

Correct Answer: A

QUESTION 92
A company is deploying an AI-powered loan approval system. The company must comply with data governance regulations for AI. Which solution will meet these requirements?

A. Modify AI outputs based on user preferences without audit trails.
B. Implement data lifecycle management to track and manage AI training data.
C. Prioritize AI inference time optimization over data residency requirements.
D. Use only synthetic data for model training to avoid compliance risks.

Correct Answer: B

QUESTION 93
A company is evaluating several large language models (LLMs) for a text summarization task. The company needs to select a metric to evaluate the quality of the summaries that the LLMs generate. Which metric will meet this requirement?

A. Recall
B. Area under the ROC curve (AUC)
C. Recall-Oriented Understudy for Gisting Evaluation (ROUGE)
D. Mean squared error (MSE)

Correct Answer: C

QUESTION 94
A company is using AI to improve its services. The company needs to ensure that the AI system is fair and explainable. The company wants to require training for members of the AI system development team. Which training will meet these requirements?

A. Training on advanced coding skills
B. Training on data privacy and encryption protocols
C. Training on bias awareness and responsible AI
D. Training on advanced ML algorithms

Correct Answer: C

QUESTION 95
A company wants to identify harmful language in the comments section of social media posts by using an ML model. The company will not use labeled data to train the model. Which strategy should the company use to identify harmful language?

A. Use Amazon Rekognition moderation.
B. Use Amazon Comprehend toxicity detection.
C. Use Amazon SageMaker AI built-in algorithms to train the model.
D. Use Amazon Polly to monitor comments.

Correct Answer: B

QUESTION 96
A company wants to use a foundation model (FM) in Amazon Bedrock to effectively handle the company’s tasks. The larger FMs in Bedrock meet the company’s needs. However, the larger FMs exceed the company’s budget constraints. The company needs a smaller FM solution that maintains adequate task performance and does not require custom training datasets. Which solution will meet these requirements?

A. Create a distilled model in Amazon Bedrock by using the current teacher models available.
B. Fine-tune a smaller model in Amazon Bedrock.
C. Perform continued pre-training on a smaller model in Amazon Bedrock.
D. Deploy multiple smaller FMs in parallel that match the capabilities of the larger FM.

Correct Answer: A

QUESTION 97
A company maintains a large product catalog with detailed descriptions. The company wants to build an AI assistant to answer customer questions about the products. However, the company’s labeled training data is limited. Which solution will meet these requirements with the LEAST implementation effort?

A. Use a Retrieval Augmented Generation (RAG) architecture to query the product database at runtime. Provide relevant unmodified product descriptions as context for the foundation model (FM).
B. Fine-tune a foundation model (FM) on the restricted labeled data. Automatically refresh the model with novel product details weekly for accuracy.
C. Deploy a foundation model (FM) for each product category. Implement a routing layer to direct customer queries to the appropriate specialized model.
D. Create a new custom foundation model (FM) that is trained on the product database. Optimize the FM for minimal token usage during inference.

Correct Answer: A

QUESTION 98
A company uses an Amazon Bedrock large language model (LLM) in an application. During testing, the company observes different outputs from the same input. What is the MOST likely cause of this issue?

A. The LLM is acting in a nondeterministic way.
B. The guardrails of the LLM are not configured properly.
C. The LLM has security vulnerabilities.
D. The LLM is acting in a deterministic way.

Correct Answer: A

QUESTION 99
A company is deploying a generative AI application on Amazon Bedrock. The company wants to restrict all application traffic to AWS so that traffic does not route to the public internet. Which solution will meet these requirements with the LEAST amount of development effort?

A. Internet gateway
B. NAT gateway
C. AWS PrivateLink
D. AWS Transit Gateway

Correct Answer: C

QUESTION 100
Which strategy will evaluate the performance of a foundation model (FM) in real-world applications?

A. Conducting AB testing with users in a controlled environment
B. Human evaluation by subject matter experts
C. Measuring the model’s accuracy on a training dataset
D. Analysis of the model’s internal representations and attention patterns

Correct Answer: A