用分层代理系统提升医疗AI安全，自动纠错率达24%。
Tiered Agentic Oversight: A Hierarchical Multi-Agent System for Healthcare Safety
- 分层代理按任务复杂度分配，模拟医护协作流程。
- 在5个医疗安全基准中4项超越单代理，最高提升8.2%。
- 实验证明下层代理不可或缺，人类医生参与可提准至60%。
将大型语言模型作为代理部署于临床环境时，因其潜在错误和单点故障而带来显著安全风险。本文提出分层代理监督（TAO），一种受医院医护层级（如护士-医师-专家）启发的分层多代理系统，通过自动化跨层与层内通信及角色扮演，实现任务路由与安全监控。该结构有效作为错误纠正机制，可吸收高达24%的个体代理错误，防止其累积。实验显示，TAO在5个医疗安全基准中的4个上优于单代理及其他多代理系统，性能最高提升8.2%。消融研究证实：(i) 动态架构比静态单层配置更安全，超出3%；(ii) 下层代理至关重要，移除后安全性能下降最明显。最后，用户研究中，医生作为最高层级代理提供修正反馈，使医疗分诊准确率从40%提升至60%。
原文摘要 · Abstract (English)
Large language models (LLMs) deployed as agents introduce significant safety risks in clinical settings due to their potential for error and single points of failure. We introduce Tiered Agentic Oversight (TAO), a hierarchical multi-agent system that enhances AI safety through layered, automated supervision. Inspired by clinical hierarchies (e.g., nurse-physician-specialist) in hospital, TAO routes tasks to specialized agents based on complexity, creating a robust safety framework through automated inter- and intra-tier communication and role-playing. Crucially, this hierarchical structure functions as an effective error-correction mechanism, absorbing up to 24% of individual agent errors before they can compound. Our experiments reveal TAO outperforms single-agent and other multi-agent systems on 4 out of 5 healthcare safety benchmarks, with up to an 8.2% improvement. Ablation studies confirm key design principles of the system: (i) its adaptive architecture is over 3% safer than static, single-tier configurations, and (ii) its lower tiers are indispensable, as their removal causes the most significant degradation in overall safety. Finally, we validated the system's synergy with human doctors in a user study where a physician, acting as the highest tier agent, provided corrective feedback that improved medical triage accuracy from 40% to 60%. Project Page: https://tiered-agentic-oversight.github.io/
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。