Free AWS Certified AI Practitioner Exam Practice Questions

Question # 41

A company wants to implement a generative AI assistant to provide consistent responses to various phrasings of user questions.

Which advantages can generative AI provide in this use case?

Options:

Low latency and high throughput

Adaptability and responsiveness

Deterministic outputs and fixed responses

Hardware acceleration and GPU optimization

Discussion 0

Question # 42

In which stage of the generative AI model lifecycle are tests performed to examine the model ' s accuracy?

Options:

Deployment

Data selection

Fine-tuning

Evaluation

Discussion 0

Question # 43

A company has an ML model. The company wants to know how the model makes predictions. Which term refers to understanding model predictions?

Options:

Model interpretability

Model training

Model interoperability

Model performance

Discussion 0

Question # 44

A company has developed a custom foundation model (FM) and wants to deploy the FM on AWS. The company needs a managed service that provides the ability to load the company ' s proprietary model weights without managing infrastructure.

Which AWS feature or resource will meet these requirements?

Options:

Amazon SageMaker JumpStart

Amazon Bedrock Custom Model Import

Amazon EC2 with AWS Deep Learning AMIs (DLAMIs)

AWS Lambda with container images

Discussion 0

Question # 45

A retail company wants to build an ML model to recommend products to customers. The company wants to build the model based on responsible practices. Which practice should the company apply when collecting data to decrease model bias?

Options:

Use data from only customers who match the demography of the company ' s overall customer base.

Collect data from customers who have a past purchase history.

Ensure that the data is balanced and collected from a diverse group.

Ensure that the data is from a publicly available dataset.

Discussion 0

Question # 46

Which use case is MOST suitable for a binary classification model?

Options:

Providing multiple product recommendations to a customer based on the customer ' s previous purchases

Determining product pricing

Detecting spam emails

Marketing products to customers based on customer demographics

Discussion 0

Question # 47

A company plans to implement a generative AI (GenAI) tool to create content that is based on trending social media topics.

Which option represents a legal risk of creating content by using this tool?

Options:

Diversity claims for making the generated content more diverse and inclusive

Bias in generated content

International trade restrictions on AI technology deployment

Discussion 0

Question # 48

Which strategy will prevent model hallucinations?

Options:

Fact-check the output of the large language model (LLM).

Compare the output of the large language model (LLM) to the results of an internet search.

Use contextual grounding.

Use relevance grounding.

Discussion 0

Question # 49

Which technique breaks a complex task into smaller subtasks that are sent sequentially to a large language model (LLM)?

Options:

One-shot prompting

Prompt chaining

Tree of thoughts

Retrieval Augmented Generation (RAG)

Discussion 0

Question # 50

A company uses a foundation model (FM) on Amazon Bedrock to generate meeting summaries and insights from discussion transcripts. However, productivity has not improved.

Which solution will help determine if the FM meets company business objectives?

Options:

Compare pre-deployment and post-deployment metrics such as time saved in documentation, number of actionable tasks created, and employee adoption rates.

Evaluate the FM’s outputs by using technical quality metrics such as precision, recall, or Bilingual Evaluation Understudy (BLEU) scores to confirm summarization accuracy.

Extend the summarization workflow with a Retrieval Augmented Generation (RAG) layer so the FM includes project notes and documents for better insights.

Review employee satisfaction surveys to understand general sentiment toward the summaries.

Discussion 0