PAAC：隐私感知的智能代理端云协作

摘要

大型语言模型（LLM）智能体面临结构性矛盾：云端智能体具备强推理能力但会暴露用户数据，而设备端智能体在保护隐私的同时却牺牲了整体能力。现有设备-云协同设计仅将此边界视为计算划分，而非适应智能体工作负载的信任边界；现有脱敏器则迫使开发者在策略灵活性与工具调用所需的结构保真度之间做出取舍。本文提出PAAC（隐私感知智能体框架），通过将规划器-执行器分解与设备-云边界对齐，使角色分工本身成为隐私保护机制。云端智能体基于类型化占位符令牌进行推理——这些令牌保留各敏感值的推理角色特征却丢弃其具体内容，而设备端智能体则识别敏感片段，将每步执行结果提炼为紧凑的关键发现。脱敏机制限定设备端LLM仅能提议需屏蔽的片段，所有替换与还原操作均由确定性注册表执行，确保动作可在设备端直接执行。在严格隐私设置下的三个智能体基准测试中，PAAC主导了隐私与准确率的帕累托前沿：相较于最先进的设备-云基线方法，平均准确率提升15-36%，平均泄露降低2-6倍，且在固定实体分类体系之外的隐私目标上降幅最大。我们在涵盖数学、科学、金融等10个领域的17个额外基准测试中观察到一致性提升。

English

Large language model (LLM) agents face a structural tension: cloud agents provide strong reasoning but expose user data, while on-device agents preserve privacy at the cost of overall capability. Existing device-cloud designs treat this boundary as a compute split rather than a trust boundary suited to agentic workloads, and existing sanitizers force a choice between policy flexibility and the structural fidelity tool calls require. In this work, we develop PAAC, a privacy-aware agentic framework that aligns planner--executor decomposition with the device-cloud boundary so that role specialization itself becomes the privacy mechanism. The cloud agent reasons over typed placeholder tokens that preserve each sensitive value's reasoning role while discarding its content, while the on-device agent identifies sensitive spans and distills each step's execution outcome into compact key findings. Sanitization confines the on-device LLM to proposing which spans to mask, while a deterministic registry performs all substitution and reversal, keeping actions directly executable on device. On three agentic benchmarks under strict privacy settings, PAAC dominates the Pareto frontier of privacy and accuracy, improving average accuracy by 15-36\% and reducing average leakage by 2-6times over state-of-the-art device-cloud baselines, with the largest margins on privacy targets outside fixed entity taxonomies. We find consistent improvements on 17 additional benchmarks spanning 10 domains, including math, science, and finance.

PAAC：隐私感知的智能代理端云协作

PAAC: Privacy-Aware Agentic Device-Cloud Collaboration

摘要

Support