Performance evaluation of GPT-4o on South Korean national exams for building mechanical equipment maintenance.
GPT-4o 在南韓建築機械設備維護國家考試的表現評估
Sci Rep 2025-08-19
Five advanced chatbots solving European Diploma in Radiology (EDiR) text-based questions: differences in performance and consistency.
五種先進聊天機器人在解決歐洲放射學文憑(EDiR)文字題的表現與一致性差異
Eur Radiol Exp 2025-08-19
What social stratifications in bias blind spot can tell us about implicit social bias in both LLMs and humans.
偏見盲點中的社會階層化對於隱性社會偏見在人類與大型語言模型(LLMs)中的啟示
Sci Rep 2025-08-19
Citation integrity in the age of AI: evaluating the risks of reference hallucination in maxillofacial literature.
AI時代的引用誠信:評估口腔顎面文獻中參考文獻幻覺的風險
J Craniomaxillofac Surg 2025-08-19
在學術寫作用 ChatGPT 這類 AI 時,常會遇到「參考文獻幻覺」問題,也就是 AI 會捏造或給錯引用,這對學術誠信很傷,特別是在重視實證的口腔顎面外科領域。建議加強編輯把關、提升 AI 素養,並用即時引用工具,確保文獻正確,維護學術公信力。
相關文章PubMedDOI推理
Assessing the transferability of BERT to patient safety: classifying multiple types of incident reports.
BERT於病人安全領域的可轉移性評估:多類型事件報告的分類
BMJ Health Care Inform 2025-08-19
Artificial intelligence in endometriosis care: A comparative analysis of large language model and human specialist responses to endometriosis-related queries.
人工智慧於子宮內膜異位症照護中的應用:大型語言模型與人類專科醫師對子宮內膜異位症相關問題回應的比較分析
Eur J Obstet Gynecol Reprod Biol 2025-08-19
這項研究發現,ChatGPT-4o 回答子宮內膜異位症相關問題的表現,和人類專家差不多,品質、正確性和安全性都不輸專家。雖然專家多能分辨 AI 回答,但 AI 仍有潛力協助病人獲取健康資訊。未來還需進一步研究 AI 在臨床應用的風險與接受度。
相關文章PubMedDOI推理