ExperienceWeaver: 优化基于 LLM 的临床文本改进的小样本经验学习
MedTextWeaver: Procedural Knowledge Evolution in Agentic Medical Text Editing
AI总结：
ExperienceWeaver 通过经验学习优化小样本下 LLM 的临床文本改进，通过提炼反馈知识提升模型修订能力。
摘要：
临床文本改进对于提高医疗效率至关重要，但仍然困难，因为高质量数据有限且医疗文档具有复杂的约束。尽管大型语言模型（LLMs）显示出潜力，但当前方法在小样本设置中面临挑战：监督微调数据密集且成本高，而检索增强生成往往只能提供表面修正，而无法捕捉修订背后的推理过程。为了解决这些限制，我们提出 ExperienceWeaver，一种分层框架，将重点从数据检索转向经验学习。与其简单地回忆过去的例子，ExperienceWeaver 将嘈杂的多维反馈提炼为结构化、可操作的知识。具体而言，是针对错误的提示和高层策略。通过将这种提炼的经验注入到代理流程中，模型学会“如何修订”而不是仅仅“是什么修订”。在四个临床数据集上的广泛评估表明，ExperienceWeaver 一致地提高了性能，在小样本设置中超越了 Gemini-3 Pro 等最先进的模型。
英文摘要：
Medical text editing is essential for improving communication among diverse stakeholders in clinical settings. However, adapting LLM agents to this task remains challenging because expert supervision is often sparse, fragmented, and distributed across interacting quality dimensions. We identify that direct accumulation or retrieval of individual feedback is insufficient for effective adaptation, as fragmented evaluations do not directly translate into a coherent understanding of medical text quality. Based on this observation, we propose MedTextWeaver, a training-free framework that transforms fragmented evaluative evidence into global quality principles and actionable procedural knowledge for medical text editing. Across three clinical text datasets and a real-world validation experiment, MedTextWeaver consistently improves performance over strong LLM baselines and existing memory-based adaptation approaches. Further analysis demonstrates that the learned knowledge enables more effective adaptation under limited supervision while providing an explicit and interpretable interface between expert evaluations and LLM editing behavior.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
