試験
通常
試験時間: 00:00問題時間: 00:00
問題一覧
問題 56 / 65
Which evaluation method is most reliable for assessing the clinical accuracy and safety of generative models in the medical field?
①Similarity evaluation using BLEU score
②Summary evaluation using ROUGE scores
③Evaluating language models using perplexity
④Peer review and validation by clinical experts
⑤Automated self-contradiction detection tool