2.3Standard
A summarization prompt was changed and responses became shorter.
Do not judge the change by length alone.
Choose two appropriate practices.
× Incorrect
Use a fixed representative and boundary set for before-and-after comparison and measure accuracy, completeness, safety, latency, and cost.
Detailed explanation
CorrectMeasure accuracy, omissions, format, safety, latency, and cost
One attractive metric can hide regressions.
IncorrectAlways rate the shortest output as best
Important content may be missing.
IncorrectReplace the evaluation data after changing the prompt
Comparability and reproducibility are lost.
CorrectUse a fixed representative and boundary evaluation set
The same inputs make the change comparable.
Check official documentation
Compare the following points with the official documentation.
AWS公式AIF-C01ガイドDomain 2.3・3.3のプロンプト評価と品質指標を確認する。
Official references
What to confirm
固定評価セットと多面的評価の必要性を説明できる。
Key points
- Fixed set
- Comparison
- Multiple metrics
Notes
- Environment: AWS公式AIF-C01試験ガイドとAWS公式ドキュメントの確認
- Compare the definition with the AIF-C01 objective and the service documentation before applying it.
- Keep the concept separate from a particular answer choice or implementation step.
Foundation review
Evaluation set
A fixed collection of representative, boundary, and risky inputs used for comparison.
Regression
A loss of previously accepted behavior after a prompt, model, or data change.
Question IDAWS-AIF-118
EnvironmentAWS公式AIF-C01試験ガイドとAWS公式ドキュメントの確認
Last technical review2026-08-20
Report an error or rights issue