跳到正文
返回论文列表
cs.MA提交于 已译

OAK-Mind:基于知识驱动的多智能体 LLM 从被动感知数据零训练推断心理健康变化

OAK-Mind: Zero-Training Mental Health Change Inference from Passive Sensing via Knowledge-Grounded Multi-Agent LLMs

Yishu Ji · Zhihan Jiang · Will Ke Wang · Yuang Fan · Anind K. Dey · Xuhai "Orson" Xu

中文摘要

可穿戴与智能手机传感器可捕捉与心理健康相关的行为信号,但传统上从这些信号推断心理健康结果依赖需要标注数据的监督模型,且对预测的解释有限。大语言模型(LLM)为基于行为证据进行推理提供了新机会,但现有基于 LLM 的被动感知方法大多仍依赖任务特定微调或传感器特定预训练。OAK-Mind 是一个零训练(zero-training)、多智能体(multi-agent)、知识驱动(knowledge-grounded)的 LLM 流水线,用于从被动感知数据推断心理健康变化。该流水线将感知数据组织为睡眠、身体活动、位置移动性和智能手机使用等行为维度,并协调基于心理健康文献的专用智能体解读纵向行为变化;随后综合各智能体评估,并应用基于问卷的症状评分,以个体基线评估为参照推断变化,同时给出基于可观察行为模式和领域知识的可读理由。在多个数据集和预测任务上,尽管没有任务特定训练,OAK-Mind 的性能与使用额外标注样本训练的最强监督基线相当或更优;在四个测试的 LLM 骨干模型上,它也较直接提示(direct prompting)表现出一致提升。消融分析支持将行为解读与症状评分结合、将分析分布到领域专用智能体以及引入领域知识的收益。研究结果凸显了知识驱动、多智能体 LLM 在从被动感知数据零训练推断心理健康变化方面的潜力。

关键要点

  1. 01现有 LLM 被动感知方法多依赖任务特定微调或传感器预训练,缺乏零训练推理方案。
  2. 02OAK-Mind 将感知数据按睡眠、活动、移动性、手机使用等维度组织,由领域知识智能体解读纵向变化。
  3. 03零训练下性能与使用额外标注样本的最强监督基线相当或更优,并在四个 LLM 骨干上优于直接提示。
  4. 04消融表明行为解读与症状评分结合、领域专用智能体分工及领域知识引入均有增益。

解读

尚无解读。

原始英文摘要

arXiv:2611.22248v1 Announce Type: cross Abstract: Wearable and smartphone sensors capture behavioral signals associated with mental health, yet inferring mental health outcomes from these signals has traditionally relied on supervised models requiring labeled data and offering limited insight into their predictions. Large language models (LLMs) offer new opportunities to reason over behavioral evidence, but existing LLM-based approaches to passive sensing have largely continued to rely on task-specific fine-tuning or sensor-specific pretraining. In this paper, we introduce \textbf{OAK-Mind}, a zer\textbf{O}-training, multi-\textbf{A}gent, \textbf{K}nowledge-grounded LLM pipeline for inferring mental health changes from passive sensing data. The pipeline organizes sensing data into behavioral dimensions including sleep, physical activity, location mobility, and smartphone use, and orchestrates specialized agents grounded in mental health literature to interpret longitudinal behavioral changes. It synthesizes their assessments and applies questionnaire-based symptom rating to infer changes relative to an individual's baseline assessment with human-readable rationale grounded in observed behavioral patterns and domain knowledge. Across datasets and prediction tasks, while without task-specific training, OAK-Mind achieves comparable or better performance than the most predictive supervised baselines trained on additional labeled samples. It also demonstrate consistent improvement upon direct prompting across four tested LLM backbones. Ablation analyses support the benefits of combining behavioral interpretation with symptom rating, distributing analysis across domain-specific agents, and incorporating domain knowledge. Our findings highlight the potential of knowledge-grounded, multi-agent LLMs for training-free inference of mental health changes from passive sensing data.

同方向论文 · cs.MA

查看全部 →