試験
通常
試験時間: 00:00問題時間: 00:00
問題一覧
問題 20 / 65
Which approaches are effective in preventing a generative model from outputting unfounded information (hallucinations)?
①Use Amazon Polly to output audio for verification.
②Use Amazon Kendra to search reliable documents and provide supporting evidence for responses.
③Ensure the accuracy of responses using only Amazon Bedrock Guardrails.
④Use Amazon Comprehend to analyze sentiment and adjust responses.
⑤Perform random validation with AWS Lambda