Using a large language model (ChatGPT) to assess risk of bias in randomized controlled trials of medical interventions: protocol for a pilot study of interrater agreement with human reviewers.

使用大型語言模型（ChatGPT）評估醫療介入隨機對照試驗的偏倚風險：與人工審查者評分一致性的前導研究計畫