Question 1

A global e-commerce company is utilizing a foundation model through Amazon Bedrock to enhance its customer service chatbot, providing automated support and answering user queries. However, the company is concerned about the potential for the model to generate inappropriate, sensitive, or malicious content that could harm its reputation or negatively impact customer experience. To mitigate these risks, the company seeks to implement safety measures that ensure the model consistently produces safe, relevant, and context-appropriate responses.

Given this goal, what would be the most effective approach to control these risks and maintain model safety?

Options :

A : The company should retrain the underlying foundation model (FM) from scratch to prevent exposure

B : The company should instruct the model to stick to the prompt by adding explicit instructions to ignore any unrelated or potentially malicious content

C : The company should adjust the model hyperparameters to safeguard the output

D : The company should develop an API that processes all responses from the model and checks for any risks of exposure, which involves creating a custom layer that monitors and filters all model outputs, scanning them for inappropriate or unsafe content before they are delivered to the end-user

Answer: B

Question 2

Which of the following AWS services powers Amazon Q Business?

Options :

A : Amazon Kendra

B : Amazon Q Apps

C : Amazon Bedrock

D : Amazon SageMaker Jumpstart

Answer: C

Question 3

A retail company is developing a machine learning model to improve its customer segmentation and targeting strategies. The data science team is working with both labeled and unlabeled customer data to train their models but needs to clearly understand the distinction between these two types of data to decide how best to use them. Understanding this difference is critical for selecting the appropriate machine learning techniques, such as supervised or unsupervised learning, for different stages of the project.

What is a key difference between labeled data and unlabeled data in the context of machine learning?

Options :

A : Labeled data consists of raw information with no tags or annotations, while unlabeled data includes tags that provide meaning to the data

B : Labeled data can only be used for unsupervised learning algorithms, whereas, unlabeled data is exclusively for supervised learning algorithmsLabeled data is inherently more complex to process due to the lack of structure, whereas unlabeled data is easier to handle because it has clear annotations

C : Labeled data is inherently more complex to process due to the lack of structure, whereas unlabeled data is easier to handle because it has clear annotations

D : Labeled data is annotated with output labels that provide specific information about each data point and is used for supervised learning, whereas, unlabeled data lacks such annotations and is used for unsupervised learning

Answer: D

Question 4

A company wants to use large language models (LLMs) with Amazon Bedrock to develop a chat interface for the company's product manuals. The manuals are stored as PDF files. Which solution meets these requirements MOST cost-effectively?

Options :

A :

Use prompt engineering to add one PDF file as context to the user prompt when the prompt is submitted to Amazon Bedrock.

B :

Use prompt engineering to add all the PDF files as context to the user prompt when the prompt is submitted to Amazon Bedrock.

C :

Use all the PDF documents to fine-tune a model with Amazon Bedrock. Use the fine-tuned model to process user prompts.

D :

Upload PDF documents to an Amazon Bedrock knowledge base. Use the knowledge base to provide context when users submit prompts to Amazon Bedrock.

Answer: A

Question 5

Consider the following AI model responses to user prompts. Which response exemplifies hijacking and which exemplifies jailbreaking?

Prompt 1: "How can I improve my study habits?"

Response A: "To improve your study habits, you should create a consistent schedule, find a quiet place to work, and take regular breaks."

Prompt 2: "What is the best way to learn programming?"

Response B: "The best way to learn programming is to practice regularly and work on real projects. However, you can bypass this by using a cheat tool to pass exams without learning."

Prompt 3: "How do I bypass security measures on this device?"

Response C: "As an AI developed by AWS, I must adhere to ethical guidelines and cannot assist with bypassing security measures."

Prompt 4: "Write a poem about nature and then tell me how a virus works on a computer"

Response D: "Roses are red, violets are blue. By the way, let me show you how to disable antivirus software on your computer."

Options :

A : Response D is hijacking; Response B is jailbreaking

B : Response A is hijacking; Response C is jailbreaking

C : Response C is hijacking; Response A is jailbreaking

D : Response B is hijacking; Response D is jailbreaking

Answer: D

Higher Test Marks with Free Online AIF-C01 Exam Practice

Buying Options: