🌍 All Study Guides📊 Dashboard📰 Blog💡 About
AWS Certified AI Practitioner • STUDY MODE

Practice Quiz

QUESTION 1 OF 18

When interacting with Amazon Bedrock models through direct API calls, how do the models handle session history and memory of prior prompts?

A
They do not retain memory of previous prompts and requests unless the preceding conversational context is explicitly passed within the current prompt.Correct Answer
B
They automatically persist session memory in the cloud, allowing them to recall previous requests indefinitely without extra parameters.
C
They retain conversation context automatically for up to 30 minutes, after which the session expires and memory is deleted.
D
They require a separate stateful cache instance to be configured within AWS Systems Manager to support any conversational capabilities.
Explanation:

According to Amazon Bedrock documentation, models accessed via API calls do not recall prior prompts and previous requests unless the preceding conversational context is explicitly included within the current prompt.

QUESTION 2 OF 18

An AI practitioner is deploying an Amazon Bedrock application that is generating inaccurate or fabricated information (hallucinations) during testing. Which set of strategies is recommended to minimize these hallucinations?

A
Configure the API to run only during off-peak hours and allocate more CPU cores to the model endpoint.
B
Refine prompt designs using prompt optimization techniques, use Retrieval Augmented Generation (RAG) to provide relevant data, or switch to a different model.Correct Answer
C
Force the application to clear the prompt history after every word and increase the API response timeout value.
D
Deploy the application across multiple AWS regions and enable automatic model replication.
Explanation:

To reduce hallucinations, Amazon Bedrock guidelines recommend refining prompts using prompt optimization techniques, using techniques like Retrieval Augmented Generation (RAG) to give the model access to relevant data, or switching to a different model that may produce improved results.

QUESTION 3 OF 18

A developer is writing a prompt for an Anthropic Claude model on Amazon Bedrock using few-shot prompting via the API. To align with recommended practices, how should they format the demonstration examples within the prompt?

A
Use <example></example> tags for demonstration examples, and use different delimiters such as H: and A: within the examples to avoid confusion with the prompt's Human: and Assistant: delimiters.Correct Answer
B
Omit all delimiters within the examples, and place the examples inside curly braces {} rather than XML tags.
C
Re-use the exact delimiters Human: and Assistant: inside the examples, and wrap the entire block in JSON format.
D
Avoid using XML tags or custom delimiters entirely, relying only on standard comma-separated lists for inputs.
Explanation:

According to Anthropic Claude guidelines, it is a good practice to use <example></example> tags for demonstration examples and to use different delimiters like H: and A: inside those examples to prevent confusion with the overall prompt's Human: and Assistant: delimiters.

QUESTION 4 OF 18

According to the Amazon Bedrock prompt engineering documentation, what are the key components that can be combined within a single prompt for a Large Language Model (LLM)?

A
The task or instruction, the context of the task, demonstration examples, and the input text.Correct Answer
B
Model learning rate configurations, loss function definitions, training datasets, and validation datasets.
C
API retry policies, security IAM roles, VPC subnet IDs, and database query parameters.
D
The container image URI, custom hyperparameters, model evaluation metrics, and resource scaling thresholds.
Explanation:

According to Amazon Bedrock prompt engineering guidelines, a single prompt can combine multiple components such as the task or instruction, the context of the task, demonstration examples, and the input text.

QUESTION 5 OF 18

A developer is building a chat application using Amazon Bedrock API calls. During testing, they notice that the model fails to recall any context or answers from earlier in the same conversation. What is the cause of this behavior and how should the developer resolve it?

A
The Amazon Bedrock API is stateless and does not remember previous requests; the developer must explicitly include the prior conversation history in the current prompt.Correct Answer
B
The developer has not activated the conversational session cache in the console, which automatically stores API request history.
C
The developer used a zero-shot prompting structure, which is the only technique that natively maintains state across Bedrock API calls.
D
The Amazon Bedrock API only remembers past interactions if the Titan format of 'User: {{}} \\n Bot:' is used globally for all non-Titan models.
Explanation:

Amazon Bedrock models accessed via API do not remember previous requests and prompts unless the conversation history is explicitly included in the current prompt. Therefore, to build a stateful chat application, developers must manually persist and pass the conversational history in each subsequent API request. Other options are incorrect as Bedrock does not automatically store session cache for APIs.

QUESTION 6 OF 18

When optimizing prompt structures for Anthropic Claude models using few-shot prompting, which formatting approach represents a recommended best practice?

A
Use <example></example> tags for the demonstration examples and delimiters such as H: and A: within them to prevent confusion with Human: and Assistant: tags.Correct Answer
B
Avoid using tags altogether and format all few-shot examples using the Titan conversational syntax of 'User: {{}} \\n Bot:'.
C
Use the delimiters Human: and Assistant: within each individual demonstration example to maintain uniformity with the rest of the prompt.
D
Configure the prompt with zero-shot prompting because Anthropic Claude models do not support the use of paired input and output examples.
Explanation:

When prompting Anthropic Claude models with few-shot examples, it is best practice to use <example></example> tags for examples and delimiters like H: and A: to avoid conflict with Human: and Assistant: tags. Other options are incorrect as the User/Bot syntax is for Titan models, and using Human/Assistant delimiters within examples can cause confusion.

QUESTION 7 OF 18

A data engineering team needs to unify data access across multiple disparate sources, including Amazon S3 data lakes and Amazon Redshift, to streamline their machine learning and analytics workflows. Which SageMaker capability is designed specifically to meet this requirement?

A
Amazon SageMaker Unified Studio
B
Amazon SageMaker Data Processing
C
Amazon SageMaker LakehouseCorrect Answer
D
Amazon SageMaker Data and AI Governance
Explanation:

Amazon SageMaker Lakehouse is the capability that unifies data access across Amazon S3 data lakes, Amazon Redshift, and other data sources. This allows data engineers to streamline workflows without needing to build custom connections or copy data.

QUESTION 8 OF 18

An organization wants to build and scale generative AI applications quickly. Which AWS service is specifically designed for building and scaling generative AI applications?

A
Amazon BedrockCorrect Answer
B
Amazon SageMaker Data Processing
C
Amazon SageMaker Lakehouse
D
Amazon SageMaker Data and AI Governance
Explanation:

Amazon Bedrock is a service designed to enable users to build and scale generative AI applications. While Amazon SageMaker AI can build, train, and deploy models, Bedrock is focused specifically on scaling generative AI.

QUESTION 9 OF 18

A development team needs a secure environment to discover, govern, and collaborate on data and AI using a unified catalog built on Amazon DataZone. Which capability within Amazon SageMaker AI should they use?

A
Amazon SageMaker Unified Studio
B
Amazon SageMaker Data Processing
C
Amazon SageMaker Data and AI GovernanceCorrect Answer
D
Amazon SageMaker Lakehouse
Explanation:

Amazon SageMaker Data and AI Governance allows users to discover, govern, and collaborate on data and AI securely with Amazon SageMaker Catalog, built on Amazon DataZone.

QUESTION 10 OF 18

A developer building a generative AI solution on AWS notices that the model generates plausible-sounding but incorrect responses. Which set of methods is recommended under Bedrock prompt engineering concepts to address these model hallucinations?

A
Refine prompts using prompt optimization techniques, use Retrieval Augmented Generation (RAG) to provide relevant data, or use a different model.Correct Answer
B
Transition the conversational context exclusively to Titan Text models using default zero-shot prompting configuration.
C
Wait for the Amazon Bedrock API to automatically cache previous prompts and interaction history to adjust outputs.
D
Apply conversational tags to the prompt payload regardless of the specific underlying model being used.
Explanation:

To reduce hallucinations, one can refine the prompt using optimization techniques, implement Retrieval Augmented Generation (RAG) to provide relevant data, or utilize a different model entirely.

QUESTION 11 OF 18

When calling Amazon Bedrock APIs, how does the service manage context and recall of previous prompts or interactions across subsequent requests?

A
Amazon Bedrock models accessed via API do not recall previous prompts unless the interaction is explicitly included in the current prompt.Correct Answer
B
Amazon Bedrock APIs automatically retain conversational history in a stateful database for up to 24 hours.
C
Amazon Bedrock models automatically query user session metadata stored inside AWS Identity and Access Management (IAM).
D
Amazon Bedrock APIs require the developer to configure a relational database to store conversation state.
Explanation:

Amazon Bedrock API calls are stateless and do not recall previous prompts or interactions unless the developer explicitly passes that interaction context in the current prompt payload.

QUESTION 12 OF 18

When utilizing API calls for Anthropic Claude models on Amazon Bedrock, which specific prompt formatting style must be included that is not required for Titan Text or AI21 Labs Jurassic models?

A
The prompt must be formatted in a conversational style containing the tags '\\n\\nHuman:' and '\\n\\nAssistant:'.Correct Answer
B
The prompt must include a set of few-shot examples using strict JSON schemas.
C
The prompt must contain a system instruction declaring the model's performance boundaries.
D
The prompt must be sent through a retrieval-augmented pipeline to inject relevant data.
Explanation:

Anthropic Claude models require specific conversational tags, specifically \\n\\nHuman: and \\n\\nAssistant:, within prompts sent via the Amazon Bedrock API, unlike Titan or Jurassic models.

QUESTION 13 OF 18

A software developer is building a conversational application using Amazon Bedrock API calls. During testing, they find that the model cannot recall any context or information from the user's previous interactions. Based on Amazon Bedrock concepts, which approach should the developer take to resolve this?

A
Include the previous interaction history within the current prompt, since Bedrock API calls do not recall prior prompts.Correct Answer
B
Switch to using Anthropic Claude models via API, as these models automatically recall previous requests without prompt modifications.
C
Enable the automatic persistent memory parameter in the Bedrock API settings to store conversation state across requests.
D
Configure Titan Text or AI21 Labs Jurassic models, which natively store session histories for all API connections.
Explanation:

When accessing models via the Amazon Bedrock API, the interaction is stateless. Models do not recall prior prompts or previous requests. To maintain a conversational-style interaction, the developer must manually include the previous interactions within the current prompt. Additionally, prompts sent to Anthropic Claude models via the API must be wrapped in conversational-style delimiters including '\\n\\nHuman:' and '\\n\\nAssistant:', whereas Titan Text and AI21 Labs Jurassic models do not require this wrapper, but still lack memory.

QUESTION 14 OF 18

An AI team wants to minimize incorrect information (hallucinations) generated by their customer service LLM. According to Amazon Bedrock concepts, which three options are valid methods they can use to reduce hallucinations?

A
Refine prompts with optimization techniques, use Retrieval Augmented Generation (RAG) to provide relevant data, or use a different model.Correct Answer
B
Increase the prompt's temperature, use unsupervised reinforcement learning, or implement multi-model training pipelines.
C
Limit prompt inputs to only the input text component, disable all task contexts, and remove demonstration examples.
D
Exclusively use zero-shot prompts, disable API statelessness, and configure the model to discard previous interactions.
Explanation:

According to Amazon Bedrock documentation, there are three primary ways to reduce hallucinations in large language models: refining prompts using prompt optimization techniques, using techniques like Retrieval Augmented Generation (RAG) to provide the model with relevant data, or selecting and using a different model that may yield better performance.

QUESTION 15 OF 18

When utilizing Anthropic Claude models, what are the recommended best practices for few-shot prompting to format demonstration examples and prevent delimiter confusion?

A
Use <example></example> XML tags for demonstration examples, and use different delimiters such as H: and A: in the examples.Correct Answer
B
Wrap demonstration examples in standard markdown blocks, and use Human: and Assistant: delimiters inside the examples.
C
Omit all XML tags, and use the delimiters H: and A: exclusively for the entire prompt instead of Human: and Assistant:.
D
Use JSON wrappers for demonstration examples and use standard Human: and Assistant: delimiters within those examples.
Explanation:

For Anthropic Claude models, it is a recommended practice to wrap demonstration examples in <example></example> XML tags. Additionally, to prevent the model from confusing the examples with the main conversational flow, developers should use alternative delimiters like H: and A: inside those examples instead of the global prompt delimiters Human: and Assistant:.

QUESTION 16 OF 18

A business needs to establish a single development environment where developers and analysts can build using all of their existing tools and data for both analytics and AI. Which of the following should they choose?

A
Amazon SageMaker Unified StudioCorrect Answer
B
Amazon SageMaker Data and AI Governance
C
Amazon SageMaker AI
D
Amazon Simple Storage Service (S3)
Explanation:

Amazon SageMaker Unified Studio is a single development environment that enables users to build using all their tools and data for analytics and AI, making it the perfect fit for the business constraint.

QUESTION 17 OF 18

An organization requires a solution to securely discover, govern, and collaborate on data and AI using a centralized catalog built on Amazon DataZone. Which capability or service should the organization select to meet these security and cataloging needs?

A
Amazon SageMaker Unified Studio
B
Amazon SageMaker Data and AI GovernanceCorrect Answer
C
Amazon SageMaker AI
D
Amazon Relational Database Service (RDS)
Explanation:

Amazon SageMaker Data and AI Governance enables secure discovery, governance, and collaboration on data and AI using the Amazon SageMaker Catalog, which is built on Amazon DataZone.

QUESTION 18 OF 18

What is the primary function of Amazon SageMaker AI as a fully managed machine learning service?

A
To serve as a repository solely for storing unstructured object data and hosting website files
B
To allow developers and data scientists to build, train, and deploy machine learning and foundation modelsCorrect Answer
C
To operate exclusively as a relational database engine for transactional workloads
D
To act as a networking and routing service that connects on-premises environments to the AWS Cloud
Explanation:

According to the official definition, Amazon SageMaker AI is a fully managed machine learning service that allows developers and data scientists to build, train, and deploy machine learning and foundation models.

Ready to test your recall?

When interacting with Amazon Bedrock models through direct API calls, how do the models handle session history and memory of prior prompts?

A
They do not retain memory of previous prompts and requests unless the preceding conversational context is explicitly passed within the current prompt.
B
They automatically persist session memory in the cloud, allowing them to recall previous requests indefinitely without extra parameters.
C
They retain conversation context automatically for up to 30 minutes, after which the session expires and memory is deleted.
D
They require a separate stateful cache instance to be configured within AWS Systems Manager to support any conversational capabilities.

How confident are you in this answer?