arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2604.01113cs.CL
收藏

CARE:具有证据不一致性的隐私合规代理推理

CARE: Privacy-Compliant Agentic Reasoning with Evidence Discordance

University of Cambridge(剑桥大学);McGill University(麦吉尔大学);MBZUAI - Mohamed bin Zayed University of Artificial Intelligence(穆罕默德·本·扎耶德人工智能大学);Columbia University(哥伦比亚大学);Mila - Quebec AI Institute(米拉-魁北克人工智能研究所)

Haochen Liu, Weien Li, Rui Song, Zeyu Li, Chun Jason Xue, Xiao-Yang Liu, Sam Nallaperuma-Herzberg, Xue Liu, Ye Yuan

AI总结:

针对医疗领域中症状与体征矛盾的问题,CARE框架通过多阶段隐私合规的代理推理,在不访问敏感数据的情况下提升冲突证据处理能力。

摘要:

大型语言模型(LLM)系统日益用于支持高风险决策,但在证据内部不一致时表现较差。本文引入MIMIC-DOS数据集,用于重症监护室(ICU)短期器官功能恶化的预测。该数据集从公开的MIMIC-IV电子健康记录数据集中构建,仅包含存在体征与症状矛盾的病例。为解决此问题,我们提出CARE:一种多阶段隐私合规的代理推理框架,远程LLM通过生成结构化类别和转换提供指导,而本地LLM利用这些类别和转换支持证据获取和最终决策。实验证明,CARE在所有关键指标上均优于多个基线设置,表明其能更稳健地处理冲突的临床证据同时保护隐私。

英文摘要:

Large language model (LLM) systems are increasingly used to support high-stakes decision-making, but they typically perform worse when the available evidence is internally inconsistent. Such a scenario exists in real-world healthcare settings, with patient-reported symptoms contradicting medical signs. To study this problem, we introduce MIMIC-DOS, a dataset for short-horizon organ dysfunction worsening prediction in the intensive care unit (ICU) setting. We derive this dataset from the widely recognized MIMIC-IV, a publicly available electronic health record dataset, and construct it exclusively from cases in which discordance between signs and symptoms exists. This setting poses a substantial challenge for existing LLM-based approaches, with single-pass LLMs and agentic pipelines often struggling to reconcile such conflicting signals. To address this problem, we propose CARE: a multi-stage privacy-compliant agentic reasoning framework in which a proprietary LLM provides guidance by generating structured categories and transitions without accessing sensitive patient data, while a local LLM uses these categories and transitions to support evidence acquisition and final decision-making. Empirically, under controlled retrospective evaluation on MIMIC-DOS, CARE achieves the best overall performance across key metrics among the evaluated LLMs and agentic workflows, showing that it can more robustly handle conflicting clinical evidence while preserving privacy.

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。

↑