Interdisciplinary Development and Fine-Tuning of CARDIO a LLM for Cardiovascular Health Education in HIV Care: A Tutorial.
跨學科開發與微調 CARDIO:用於 HIV 照護中心血管健康教育的 LLM 教學指南
J Med Internet Res 2025-08-12
Demystifying Chains, Trees, and Graphs of Thoughts.
解密 Chains、Trees 與 Graphs of Thoughts
IEEE Trans Pattern Anal Mach Intell 2025-08-12
這篇論文回顧了用結構化提示(像是Chain-of-Thought、Tree of Thoughts等)來提升大型語言模型推理能力的最新進展,並提出一套通用框架和分類法,分析不同結構和演算法對效能的影響。作者也比較現有方法、討論理論基礎,並指出未來研究方向,期望推動更有效的提示工程策略。
相關文章PubMedDOI推理
Ethical Awareness in the Use of Large Language Models: Development and Validation of a Scale for Healthcare Professionals.
大型語言模型應用於醫療的倫理意識:醫療專業人員量表的開發與驗證
J Eval Clin Pract 2025-08-12
One Year On: Assessing Progress of Multimodal Large Language Model Performance on RSNA 2024 Case of the Day Questions.
一年後:評估多模態大型語言模型在 RSNA 2024 每日病例問題上的表現進展
Radiology 2025-08-12
最新研究發現,OpenAI 的多模態大型語言模型(如 GPT-4o)在放射學測驗的表現大幅進步,準確率幾乎追上專業放射科醫師,並且明顯優於 Google Gemini 和 Meta Llama。OpenAI 最佳模型和醫師的表現差異,統計上已無顯著差別。
相關文章PubMedDOI推理
Rethinking survey development in health research with AI-driven methodologies.
以 AI 驅動方法重新思考健康研究中的問卷開發
Front Digit Health 2025-08-12
Evaluating Artificial Intelligence's Role in Developing Research Questions in Head and Neck Reconstruction.
人工智慧在頭頸重建研究問題發展中的角色評估
Plast Reconstr Surg Glob Open 2025-08-12
這項研究發現,ChatGPT 和 Gemini 兩款 AI 都能有效協助產生頭頸重建領域系統性回顧的新研究想法,且有七成是全新未見於文獻。兩者表現沒有明顯差異,但提出的想法多偏向較廣泛的主題。
相關文章PubMedDOI推理
Combining Real and Synthetic Data to Overcome Limited Training Datasets in Multimodal Learning.
結合真實與合成資料以克服多模態學習中訓練資料集有限的問題
medRxiv 2025-08-12