Large language models provide discordant information compared to ophthalmology guidelines.
大型語言模型提供的資訊與眼科指引不一致
Sci Rep 2025-07-02
這項研究比較五款大型語言模型在回答美國眼科指引問題時的表現,大多數模型答案和指引一致率超過八成,但偶爾會出現錯誤或有害建議。Gemini 1.5 Pro 資料來源最透明,但也最常給出潛在有害答案。總結來說,LLMs 在眼科領域有幫助,但建議仍需專業審核。
相關文章PubMedDOI推理
A fake news detection model using the integration of multimodal attention mechanism and residual convolutional network.
結合多模態注意力機制與殘差卷積網路的假新聞偵測模型
Sci Rep 2025-07-02
Exploring the potential of lightweight large language models for AI-based mental health counselling task: a novel comparative study.
輕量級大型語言模型於AI心理健康諮詢任務之潛力探索:一項新穎的比較研究
Sci Rep 2025-07-02
A methodology for designing accurate, modifiable and reproducible scientific graphics in environmental studies using GPT4Designer.
在環境研究中,使用 GPT4Designer 設計精確、可修改及可重現科學圖表的方法學
Sci Rep 2025-07-02
A publicly available benchmark for assessing large language models' ability to predict how humans balance self-interest and the interest of others.
一個用於評估大型語言模型預測人類如何平衡自我利益與他人利益能力的公開基準
Sci Rep 2025-07-02