Amazon AWS Certified AI Practitioner AIF-C01 Practice Test 40
Question 390
A team wants to compare how much energy and infrastructure a larger foundation model consumes relative to a smaller suitable model. A colleague suggests: Assume the largest model always has the best environmental outcome. Which finding BEST explains why the suggestion should be rejected?
- Shifting populations can invalidate historical quality conclusions
- Model scale can increase compute demand without proportional task benefit
- The shortcut is safe because its terminology is familiar to the team
- Encryption protects data and does not explain a model's decisions
Explanation
Correct Answer
B:Model scale can increase compute demand without proportional task benefit. This is the direct defect in the proposed action: Assume the largest model always has the best environmental outcome.
Incorrect Answers
A:This concern belongs to another situation: A classification system performs well in testing but becomes less accurate after production data patterns shift.
C:Familiar wording is not proof that an action is appropriate. The actual defect is: Model scale can increase compute demand without proportional task benefit.
D:This finding is associated with a different problem: A credit-risk model's feature contributions are opaque and reviewers want to examine potentially discriminatory input influence.
Question 391
A team wants to compare how much energy and infrastructure a larger foundation model consumes relative to a smaller suitable model. After implementing Include sustainability and resource use in model selection, which observation BEST verifies the intended result?
- Bias and feature-attribution outputs support a documented review
- Model selection documents quality and compute tradeoffs for tested workloads
- The team schedules a future test of whether model selection documents quality and compute tradeoffs for tested workloads, but has no result
- The team approves Include sustainability and resource use in model selection but does not run an operational test
Explanation
Correct Answer
B:The observed result, Model selection documents quality and compute tradeoffs for tested workloads, provides direct evidence. Responsible AI considers resource efficiency alongside capability and cost.
Incorrect Answers
A:This observation would be evidence for another case: A credit-risk model's feature contributions are opaque and reviewers want to examine potentially discriminatory input influence, not the current scenario.
C:Verify the actual outcome: Model selection documents quality and compute tradeoffs for tested workloads.
D:The required observable condition is: Model selection documents quality and compute tradeoffs for tested workloads.
Question 392
The initial scenario was: A team wants to compare how much energy and infrastructure a larger foundation model consumes relative to a smaller suitable model. The original control is already in place. Now a small model performs poorly on the most important tasks. Which ADDITIONAL action BEST addresses the new condition?
- Rebalance sustainability against demonstrated quality and safety requirements
- Model selection documents quality and compute tradeoffs for tested workloads
- Assume the largest model always has the best environmental outcome
- Include sustainability and resource use in model selection
Explanation
Correct Answer
A:A small model performs poorly on the most important tasks. The appropriate added action is: Rebalance sustainability against demonstrated quality and safety requirements.
Incorrect Answers
B:This describes evidence for the original approach, not a new action: Model selection documents quality and compute tradeoffs for tested workloads.
C:This original shortcut remains unsuitable: Model scale can increase compute demand without proportional task benefit.
D:This was the original control and is already implemented. Responsible AI considers resource efficiency alongside capability and cost.
Question 393
A credit-risk model's feature contributions are opaque and reviewers want to examine potentially discriminatory input influence. Which option BEST meets the stated AI/ML or AWS requirement?
- Treat storage encryption as a direct explanation of prediction features
- Prepare documented model information and limitations
- Use Amazon SageMaker Clarify for applicable bias and explainability analysis
- Monitor drift and quality with appropriate observability tools such as SageMaker Model Monitor
Explanation
Correct Answer
C:Clarify supports analysis of bias and feature attribution in ML workflows. In this situation, choose: Use Amazon SageMaker Clarify for applicable bias and explainability analysis.
Incorrect Answers
A:Encryption protects data and does not explain a model's decisions. The proposal, Treat storage encryption as a direct explanation of prediction features, does not satisfy this need.
B:This is appropriate only after a distinct change: Stakeholders later need a concise summary of the model's intended uses. It is not the best initial response to a credit-risk model's feature contributions are opaque and reviewers want to examine potentially discriminatory input influence.
D:This can be valid for a different task: A classification system performs well in testing but becomes less accurate after production data patterns shift. The current request is different and needs the selected approach.
Question 394
A credit-risk model's feature contributions are opaque and reviewers want to examine potentially discriminatory input influence. A colleague suggests: Treat storage encryption as a direct explanation of prediction features. Which finding BEST explains why the suggestion should be rejected?
- Encryption protects data and does not explain a model's decisions
- Shifting populations can invalidate historical quality conclusions
- Overall predictive accuracy does not establish fairness across groups
- The shortcut is safe because its terminology is familiar to the team
Explanation
Correct Answer
A:Encryption protects data and does not explain a model's decisions. This is the direct defect in the proposed action: Treat storage encryption as a direct explanation of prediction features.
Incorrect Answers
B:This finding is associated with a different problem: A classification system performs well in testing but becomes less accurate after production data patterns shift.
C:This concern belongs to another situation: A hiring-screening model performs well overall but rejects qualified candidates from one demographic subgroup more often.
D:Familiar wording is not proof that an action is appropriate. The actual defect is: Encryption protects data and does not explain a model's decisions.
Question 395
A credit-risk model's feature contributions are opaque and reviewers want to examine potentially discriminatory input influence. After implementing Use Amazon SageMaker Clarify for applicable bias and explainability analysis, which observation BEST verifies the intended result?
- Bias and feature-attribution outputs support a documented review
- Monitoring detects material input or output deviations and triggers review
- The team schedules a future test of whether bias and feature-attribution outputs support a documented review, but has no result
- The team approves Use Amazon SageMaker Clarify for applicable bias and explainability analysis but does not run an operational test
Explanation
Correct Answer
A:The observed result, Bias and feature-attribution outputs support a documented review, provides direct evidence. Clarify supports analysis of bias and feature attribution in ML workflows.
Incorrect Answers
B:This observation would be evidence for another case: A classification system performs well in testing but becomes less accurate after production data patterns shift, not the current scenario.
C:An unperformed future check is not evidence. Verify the actual outcome: Bias and feature-attribution outputs support a documented review.
D:The required observable condition is: Bias and feature-attribution outputs support a documented review.
Question 396
The initial scenario was: A credit-risk model's feature contributions are opaque and reviewers want to examine potentially discriminatory input influence. The original control is already in place. Now stakeholders later need a concise summary of the model's intended uses. Which ADDITIONAL action BEST addresses the new condition?
- Prepare documented model information and limitations
- Treat storage encryption as a direct explanation of prediction features
- Use Amazon SageMaker Clarify for applicable bias and explainability analysis
- Bias and feature-attribution outputs support a documented review
Explanation
Correct Answer
A:Stakeholders later need a concise summary of the model's intended uses. The appropriate added action is: Prepare documented model information and limitations.
Incorrect Answers
B:This original shortcut remains unsuitable: Encryption protects data and does not explain a model's decisions.
C:This was the original control and is already implemented. Clarify supports analysis of bias and feature attribution in ML workflows.
D:This describes evidence for the original approach, not a new action: Bias and feature-attribution outputs support a documented review.
Question 397
A classification system performs well in testing but becomes less accurate after production data patterns shift. Which THREE statements together specify the correct approach, its rationale and direct evidence of operation? (Select THREE.)
- Evaluate outcomes by relevant subgroup and investigate potential bias
- Monitoring detects material input or output deviations and triggers review
- Monitor drift and quality with appropriate observability tools such as SageMaker Model Monitor
- The solution is confirmed merely because Monitor drift and quality with appropriate observability tools such as SageMaker Model Monitor was approved on paper
- Freeze the original evaluation and assume production conditions never change
- Continuous monitoring helps identify degradation after deployment
Explanation
Correct Answers
B:Correct verification. Observing monitoring detects material input or output deviations and triggers review is operational evidence of the desired outcome.
C:Correct approach for this case. Continuous monitoring helps identify degradation after deployment. It addresses the stated business requirement.
F:Correct reasoning. This explicitly explains why the recommended action, Monitor drift and quality with appropriate observability tools such as SageMaker Model Monitor, meets the described requirement.
Incorrect Answers
A:This control applies instead to A hiring-screening model performs well overall but rejects qualified candidates from one demographic subgroup more often. It is not a direct answer to the current scenario.
D:Approval of an action without testing is not evidence of effectiveness. Verify: Monitoring detects material input or output deviations and triggers review.
E:Incorrect action: Shifting populations can invalidate historical quality conclusions. It does not meet the present requirement.
Question 398
A classification system performs well in testing but becomes less accurate after production data patterns shift. A colleague suggests: Freeze the original evaluation and assume production conditions never change. Which finding BEST explains why the suggestion should be rejected?
- The shortcut is safe because its terminology is familiar to the team
- Overall predictive accuracy does not establish fairness across groups
- Record count alone does not ensure demographic or contextual diversity
- Shifting populations can invalidate historical quality conclusions
Explanation
Correct Answer
D:Shifting populations can invalidate historical quality conclusions. This is the direct defect in the proposed action: Freeze the original evaluation and assume production conditions never change.
Incorrect Answers
A:Familiar wording is not proof that an action is appropriate. The actual defect is: Shifting populations can invalidate historical quality conclusions.
B:This finding is associated with a different problem: A hiring-screening model performs well overall but rejects qualified candidates from one demographic subgroup more often.
C:This concern belongs to another situation: An organization curates a training dataset from customers in only one large city while serving rural communities too.
Question 399
A classification system performs well in testing but becomes less accurate after production data patterns shift. After implementing Monitor drift and quality with appropriate observability tools such as SageMaker Model Monitor, which observation BEST verifies the intended result?
- The team approves Monitor drift and quality with appropriate observability tools such as SageMaker Model Monitor but does not run an operational test
- The team schedules a future test of whether monitoring detects material input or output deviations and triggers review, but has no result
- Monitoring detects material input or output deviations and triggers review
- Subgroup evaluations quantify disparities and inform corrective review
Explanation
Correct Answer
C:The observed result, Monitoring detects material input or output deviations and triggers review, provides direct evidence. Continuous monitoring helps identify degradation after deployment.
Incorrect Answers
A:The required observable condition is: Monitoring detects material input or output deviations and triggers review.
B:Verify the actual outcome: Monitoring detects material input or output deviations and triggers review.
D:This observation would be evidence for another case: A hiring-screening model performs well overall but rejects qualified candidates from one demographic subgroup more often, not the current scenario.
Question 400
The initial scenario was: A classification system performs well in testing but becomes less accurate after production data patterns shift. The original control is already in place. Now the business later receives verified correction labels from users. Which ADDITIONAL action BEST addresses the new condition?
- Monitor drift and quality with appropriate observability tools such as SageMaker Model Monitor
- Freeze the original evaluation and assume production conditions never change
- Monitoring detects material input or output deviations and triggers review
- Use them in a governed re-evaluation and possible retraining workflow
Explanation
Correct Answer
D:The business later receives verified correction labels from users. The appropriate added action is: Use them in a governed re-evaluation and possible retraining workflow.
Incorrect Answers
A:This was the original control and is already implemented. Continuous monitoring helps identify degradation after deployment.
B:This original shortcut remains unsuitable: Shifting populations can invalidate historical quality conclusions.
C:This describes evidence for the original approach, not a new action: Monitoring detects material input or output deviations and triggers review.