面向可穿戴数据的个性化睡眠指导：基于语言模型的方法
Toward Personalized Sleep Guidance from Wearable Data Using Language Models
浏览论文内容
中文总结 AI 辅助
针对可穿戴数据睡眠指导不足，提出两阶段框架：先用多智能体LLM构建数据集，再蒸馏至小模型并采用Best-of-N策略，实验证明其优于商业及开源模型。
中文摘要 AI 辅助
利用可穿戴数据进行睡眠监测在个人健康领域展现出前景，然而，基于大型语言模型（LLM）的摘要生成和问答仍不足以提供个性化的睡眠指导。训练专用模型通常需要昂贵的专家标注。此外，隐私和可访问性问题促使为终端用户提供轻量级、本地化部署。我们提出一个两阶段框架来解决这些挑战。具体而言，在第一阶段，一个多智能体LLM流水线从未标注的可穿戴记录中推理出结构化的睡眠指导，从而实现可扩展的数据集构建。第二阶段通过监督微调将指导推理轨迹提炼到小型语言模型（SLM）中，并整合一种无需训练的Best-of-N选择策略以增强推理能力。实验结果表明，我们的方法优于商业通用和医疗LLM以及开源模型。人工评估进一步支持所生成指导的质量以及使用SLM进行个性化睡眠指导的可行性。
英文摘要
Sleep monitoring using wearable data has shown promise for personal health, yet large language model (LLM)-based summarization and question answering remain insufficient for personalized sleep guidance. Training specialized models, however, often requires costly expert annotation. Moreover, privacy and accessibility concerns motivate lightweight, local deployment for end users. We present a two-stage framework to address these challenges. Specifically, in Stage~1, a multi-agent LLM pipeline reasons structured sleep guidance from unannotated wearable records, enabling scalable dataset construction. Stage~2 distills guidance reasoning trajectories into small language models (SLMs) through supervised fine-tuning and integrates a training-free Best-of-$N$ selection strategy to enhance inference. Experimental results demonstrate our method outperforms commercial general and medical LLMs and open-source models. Human evaluation further supports the quality of the generated guidance and the feasibility of personalized sleep guidance with SLMs.
发表机构
- The University of Chicago（芝加哥大学）
- Baylor College of Medicine（贝勒医学院）
- Georgetown University School of Medicine（乔治城大学医学院）
机构由 AI 辅助整理，请以论文原文为准。
