AWS Certification Exam AI Practitioner (Practice Questions) set3
通常
試験時間: 00:00問題時間: 00:00
問題一覧
問題 3 / 65
I want to reduce the risk that an AI's responses contain biased or inappropriate language. What is most appropriate in prompt engineering?
①Train by adding data randomly.
②Simply retrain on a larger dataset.
③Try to reduce risk simply by shortening the length of the output.
④List only the prohibited words in the negative prompt.
⑤Include safety policies and concrete examples of disallowed expressions in the prompt and combine censorship and moderation procedures.