PAAC：隱私感知自主裝置-雲端協作

摘要

大語言模型（LLM）代理面臨一項結構性矛盾：雲端代理雖具備強大推理能力，卻會暴露使用者資料；而裝置端代理雖能保護隱私，卻需以整體能力為代價。現有的裝置-雲端協作設計僅將此邊界視為計算劃分，而非適合代理工作負載的信任邊界；現有的清理機制則迫使使用者在策略靈活性與工具呼叫所需的結構保真度之間做出取捨。本研究提出PAAC，這是一個隱私感知的代理框架，將規劃器-執行器分解機制與裝置-雲端邊界對齊，使角色專業化本身成為隱私保護機制。雲端代理基於類型化佔位符令牌進行推理，這些令牌保留每個敏感值的推理角色，同時捨棄其具體內容；裝置端代理則辨識敏感片段，並將每個步驟的執行結果提煉為精簡的關鍵發現。清理機制將裝置端LLM的權限限制於提出應遮蔽的片段，而由確定性登錄表負責所有替換與還原操作，使各行動能直接在裝置端執行。在嚴格隱私設定下的三個代理基準測試中，PAAC主導了隱私與準確度的帕累托前沿：相較於最先進的裝置-雲端基線，平均準確度提升15%至36%，平均洩漏量降低2至6倍，且對於固定實體分類體系之外的隱私目標，提升幅度最為顯著。我們在橫跨數學、科學與金融等10個領域的17個額外基準測試中，亦觀察到一致的改善成效。

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：隱私感知自主裝置-雲端協作

PAAC: Privacy-Aware Agentic Device-Cloud Collaboration

摘要

Support