Comparing scientific abstracts generated by ChatGPT to real abstracts with detectors and blinded human reviewers
Catherine A. Gao, Frederick Matthew Howard, Nikolay S. Markov, Emma Dyer, Siddhi Ramesh, Yuan Luo, Alexander T. Pearson
Northwestern University University of Chicago
阅读操作
确认中在文库中上传 PDF 后可生成中文音频讲解。
摘要与影响
Large language models such as ChatGPT can produce increasingly realistic text, with unknown information on the accuracy and integrity of using these models in scientific writing. We gathered fifth research abstracts from five high-impact factor medical journals and asked ChatGPT to generate research abstracts based on their titles and journals. Most generated abstracts were detected using an AI output detector, 'GPT-2 Output Detector', with % 'fake' scores (higher meaning more likely to be generated) of median [interquartile range] of 99.98% 'fake' [12.73%, 99.98%] compared with median 0.02% [IQR 0.02%, 0.09%] for the original abstracts. The AUROC of the AI output detector was 0.94. Generated abstracts scored lower than original abstracts when run through a plagiarism detector website and iThenticate (higher scores meaning more matching text found). When given a mixture of original and general abstracts, blinded human reviewers correctly identified 68% of generated abstracts as being generated by ChatGPT, but incorrectly identified 14% of original abstracts as being generated. Reviewers indicated that it was surprisingly difficult to differentiate between the two, though abstracts they suspected were generated were vaguer and more formulaic. ChatGPT writes believable scientific abstracts, though with completely generated data. Depending on publisher-specific guidelines, AI output detectors may serve as an editorial tool to help maintain scientific standards. The boundaries of ethical and acceptable use of large language models to help scientific writing are still being discussed, and different journals and conferences are adopting varying policies.
逐年被引趋势
关键指标
同类平均 = 1
同领域 · 同年份 · 同类型
Google Scholar 与 OpenAlex 的被引统计范围不同,数值存在差异属正常。
AI 辅助阅读
依据:摘要
可就本文提问;依据不足时会说明。
学术脉络
学科主题
生物医学Artificial Intelligence in Healthcare and Education
Explainable Artificial Intelligence (XAI) · Topic Modeling
参考文献 20
此处列出前 3 条
引用本文 715
按被引量排序,此处列出前 3 条