3.1Standard
Move a large model's behavior to a low-latency small model.
Evaluate quality and training-data risk.
Choose two considerations.
× Incorrect
Evaluate the small model's quality, safety, format, latency, and cost and govern teacher outputs used as adaptation data.
Detailed explanation
IncorrectUse all teacher outputs without checking
Errors and sensitive content can be inherited.
CorrectCheck rights, confidentiality, and errors in teacher outputs
Generated training examples can contain bias or secrets.
CorrectEvaluate quality, harmful inputs, format, latency, and cost of the small model
A smaller model may not preserve every capability.
IncorrectA small model always has the same reasoning ability
Capabilities can differ.
Check official documentation
Compare the following points with the official documentation.
AWS公式AIF-C01 Domain 3.1のモデル適応・評価と責任あるAIを確認する。
Official references
What to confirm
蒸留の利点と、能力・安全性・データ品質の再評価を説明できる。
Key points
- Small model
- Distillation data
- Reevaluation
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
Knowledge distillation
Transferring behavior or representations from a larger teacher model to a smaller student model.
Teacher output
A response or score produced by a source model and reused as data for another model.
Question IDAWS-AIF-181
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