研究人员开发了新的基准和框架，以评估和改进大型语言模型（LLM）在临床环境中的性能。PhysicianBench 为真实世界电子健康记录（EHR）任务上的 LLM 代理提供了全面的评估，显示出当前成功率低于 50% 的局限性。此外，ReMedi 提供了一个框架，通过生成改进的理由-答案对进行微调，以增强从 EHR 中预测临床结果。另一种方法引入了一种轻量级的检索增强生成方法，用于可扩展的患者-试验匹配，以降低计算成本实现了与端到端 LLM 方法相当的性能。 AI
影响 这些进展旨在提高 LLM 在医疗保健领域的准确性和效率，可能带来更好的患者护理和试验匹配。
排序理由 多篇研究论文介绍了用于评估和改进 LLM 在临床环境中性能的新基准和框架。
AI 生成摘要 · Google Gemini · 来自 5 个来源。 我们如何撰写摘要 →
arXiv:2512.20822v2 Announce Type: replace Abstract: Large Language Models (LLMs) are increasingly applied to medicine, yet their adoption is limited by concerns over reliability and safety. Existing evaluations either test factual medical knowledge in isolation or assess patient-…
arXiv:2605.02240v1 Announce Type: new Abstract: We introduce PhysicianBench, a benchmark for evaluating LLM agents on physician tasks grounded in real clinical setting within electronic health record (EHR) environments. Existing medical agent benchmarks primarily focus on static …
arXiv:2605.01474v1 Announce Type: new Abstract: Predicting future clinical outcomes from electronic health records (EHR) remains challenging due to the complexity and heterogeneity of patient data. LLMs have shown strong potential for such predictive tasks, yet existing approache…
arXiv:2604.22061v1 Announce Type: new Abstract: Patient-trial matching requires reasoning over long, heterogeneous electronic health records (EHRs) and complex eligibility criteria, posing significant challenges for scalability, generalization, and computational efficiency. Exist…
Patient-trial matching requires reasoning over long, heterogeneous electronic health records (EHRs) and complex eligibility criteria, posing significant challenges for scalability, generalization, and computational efficiency. Existing approaches either rely on full-document proc…