提出动态调整代理任务委派安全性的框架，让系统随上下文变化自动权衡安全与效率。
Safe Bilevel Delegation (SBD): A Formal Framework for Runtime Delegation Safety in Multi-Agent Systems
- 将委派问题建模为双层优化，外层学习上下文相关的安全权重
- 理论证明安全单调性、策略收敛性及责任传递上限，确保可问责性
- 在医疗、金融、教育三领域验证框架可行性，支持高风险场景应用
随着大语言模型代理部署于高风险环境，如何安全地将子任务委派给专业化子代理成为关键问题。现有工作仅在设计阶段选择多代理架构或提供宽泛的经验指导，缺乏运行时根据任务上下文动态调节安全-效率平衡的机制。本文提出安全双层委派（SBD）框架，用于层次化多代理系统的运行时委派安全。SBD将任务委派建模为双层优化：外层元权重网络 φ 学习上下文依赖的安全-效率权重 λ(s) ∈ [0,1]；内层优化委派策略 π，满足概率安全约束 P(安全) ≥ 1−δ。连续的委派度 α ∈ [0,1] 控制向每个子代理转移的决策权，平滑介于完全人工干预（α=0）与完全自主执行（α=1）之间。本文建立三项理论结果：（1）安全单调性——更高外层安全权重产生更安全的内层策略；（2）内层策略收敛性——在标准光滑性假设下，投影梯度下降线性收敛；（3）责任传播界，为多跳委派链中每代理分配可证明的责任上限。我们在三个高风险领域实例化 SBD：医疗人工智能（MIMIC-III）、金融风控（S&P 500）和教育代理监督（ASSISTments），明确数据集、安全约束集、基线和评估协议。本文完整呈现形式框架与理论结果；基于所述协议的实证验证将在后续修订中报告。
原文摘要 · Abstract (English)
As large language model (LLM) agents are deployed in high-stakes environments, the question of how safely to delegate subtasks to specialized sub-agents becomes critical. Existing work addresses multi-agent architecture selection at design time or provides broad empirical guidelines, but neither provides a runtime mechanism that dynamically adjusts the safety-efficiency trade-off as task context changes during execution. We propose Safe Bilevel Delegation (SBD), a formal framework for runtime delegation safety in hierarchical multi-agent systems. SBD formulates task delegation as a bilevel optimization problem: an outer meta-weight network phi learns context-dependent safety-efficiency weights lambda(s) in [0,1]; an inner loop optimizes the delegation policy pi subject to a probabilistic safety constraint P(safe) >= 1-delta. The continuous delegation degree alpha in [0, 1] controls how much decision authority is transferred to each sub-agent, interpolating smoothly between full human override (alpha=0) and fully autonomous execution (alpha=1). We establish three theoretical results: (1) Safety Monotonicity--higher outer safety weight produces a weakly safer inner policy; (2) Inner Policy Convergence--projected gradient descent on the inner problem converges linearly under standard smoothness assumptions; (3) an Accountability Propagation bound that distributes responsibility across multi-hop delegation chains with a provable per-agent ceiling. We instantiate SBD in three high-stakes domains--medical AI (MIMIC-III), financial risk control (S and P 500), and educational agent supervision (ASSISTments)--specifying datasets, safety constraint sets, baselines, and evaluation protocols. This manuscript presents the formal framework and theoretical results in full; empirical validation following the protocols described herein is planned and will be reported in a forthcoming revision.
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。