AWS Certification Exam AI Practitioner (Practice Questions) set14
通常
試験時間: 00:00問題時間: 00:00
問題一覧
問題 38 / 65
What is the most direct measure to prevent prompt tampering (prompt injection) caused by malicious input from external sources?
①Provide additional few-shot examples.
②Only carry out post-checks of the output results.
③Always enable Chain-of-Thought.
④Set the temperature to 0 to make it deterministic.
⑤Sanitize and validate input.