arXiv cs.MA· Hei Ting Chan, Chenwei Wu, Xueshen Liu, Boyuan Zheng, Liyue Shen, Jiasi Chen, Z. Morley Mao·· 12 小时前AI 评分38
MedCache：面向纵向临床智能体的高效时序有效记忆框架
MedCache: Efficient and Temporally Valid Memory for Longitudinal Clinical Agents
AI 导读
MedCache 是一个面向纵向临床智能体的混合记忆框架，通过构建时序有效的患者记忆、将证据组织为重叠的专科视图、按查询路由相关记忆并自适应调用单个或多个专科智能体，在推理准确率和记忆效率上优于强单智能体与多智能体基线。
正文
Abstract:Longitudinal clinical agents must maintain an evolving patient state from evidence distributed across visits, time points, and specialties. However, how agent memory should be designed for this setting remains unclear. We introduce a benchmark of multi-visit, multi-specialty patient records that evaluates long-context evidence retrieval, cross-time evidence aggregation, and cross-specialty clinical reasoning. Using this benchmark, we systematically study four memory design choices: curation, organization, retrieval, and memory-augmented reasoning. We find that temporal validity is more important than simply retaining more history; specialty-factorized memory reduces context but can hide shared evidence; and multiple agents help when specialists must reason together, not merely when evidence comes from multiple memories. Guided by these findings, we propose \textit{MedCache}, a hybrid framework that constructs temporally valid patient memory, organizes evidence into overlapping specialty views, routes each query to relevant memories, and adaptively invokes one or multiple specialists. Experiments show that MedCache improves reasoning accuracy and memory efficiency over strong single-agent and multi-agent baselines, while generalizing across model backbones and external datasets.
| Subjects: | Machine Learning (cs.LG); Distributed, Parallel, and Cluster Computing (cs.DC); Multiagent Systems (cs.MA) |
| Cite as: | arXiv:2608.29528 [cs.LG] |
| (or arXiv:2608.29528v2 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2608.29528 arXiv-issued DOI via DataCite |
Submission history
From: Hei Ting Chan [view email]
[v1]
Sun, 30 Aug 2026 03:28:56 UTC (851 KB)
[v2]
Sat, 24 Oct 2026 05:13:55 UTC (851 KB)
来源：arXiv cs.MA · arxiv.org