Practice Quiz
A financial organization is building a machine learning model to assess loan eligibility. The development team wants to mitigate bias by examining specific attributes at multiple stages, including during data preparation, after model training, and within their deployed models. Which AWS service should they use to detect and mitigate potential bias across these stages?
According to the AWS documentation, Amazon SageMaker Clarify helps mitigate bias by examining specific attributes to detect potential bias during data preparation, after model training, and within deployed models. Amazon SageMaker Model Monitor is used to automatically detect and alert on inaccurate predictions from deployed models, Bedrock Guardrails provides safeguards to filter content, and AI Service Cards act as a transparency resource documenting use cases and limitations.
Verified References
An organization is looking to implement a consistent safety and privacy solution across multiple foundation models, including models hosted in Amazon Bedrock as well as self-hosted models. They require configurable safeguards to filter harmful text and image content, redact sensitive information, and detect model hallucinations. Which AWS capability meets these requirements?
Amazon Bedrock Guardrails provides configurable safeguards to detect and filter harmful text and image content, redact sensitive information, and detect model hallucinations across any foundation model in Amazon Bedrock or self-hosted setups. Amazon SageMaker Model Monitor detects inaccurate predictions, AWS AI Service Cards provide transparency, and SageMaker Clarify is used for bias detection.
Verified References
According to the AWS framework for responsible AI, which of the following represents the core set of dimensions defined by AWS to assess and update responsible AI practices?
AWS defines responsible AI using a core set of dimensions that includes Fairness, Explainability, Privacy and security, Safety, Controllability, Veracity and robustness, Governance, and Transparency.
Verified References
An organization wants to improve transparency for its AI services and models. Which AWS resource should they refer to for a single, consolidated document detailing intended use cases, limitations, responsible AI design choices, and performance optimization best practices?
AWS AI Service Cards provide a single resource detailing intended use cases, limitations, responsible AI design choices, and performance optimization best practices to enhance transparency for AI services and models.
Verified References
A security team needs to implement customized safeguards for their generative AI applications across both Amazon Bedrock models and self-hosted foundation models. They require a solution that can automatically redact sensitive information, detect model hallucinations, and filter out harmful text and image content. Which AWS capability should they configure?
Amazon Bedrock Guardrails enables the implementation of customized safeguards to detect and filter harmful text and image content, redact sensitive information, and detect model hallucinations across any foundation model on Bedrock or self-hosted.
Verified References
Ready to test your recall?
A financial organization is building a machine learning model to assess loan eligibility. The development team wants to mitigate bias by examining specific attributes at multiple stages, including during data preparation, after model training, and within their deployed models. Which AWS service should they use to detect and mitigate potential bias across these stages?
How confident are you in this answer?