Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2604.10145 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908954874871808 |
|---|---|
| author | Li, Xiaodong Wang, Yuhua Yu, Qingchen Qin, Zixuan Sun, Yifan Zhang, Qinnan Zhang, Hainan Zheng, Zhiming |
| author_facet | Li, Xiaodong Wang, Yuhua Yu, Qingchen Qin, Zixuan Sun, Yifan Zhang, Qinnan Zhang, Hainan Zheng, Zhiming |
| contents | Client-side privacy rewriting is crucial for deploying LLMs in privacy-sensitive domains. However, existing approaches struggle to balance privacy and utility. Full-text methods often distort context, while span-level approaches rely on impractical manual masks or brittle static dictionaries. Attempts to automate localization via prompt-based LLMs prove unreliable, as they suffer from unstable instruction following that leads to privacy leakage and excessive context scrubbing. To address these limitations, we propose DAMPER (Domain-Aware Mask-free Privacy Extraction and Rewriting). DAMPER operationalizes latent privacy semantics into compact Domain Privacy Prototypes via contrastive learning, enabling precise, autonomous span localization. Furthermore, we introduce a Prototype-Guided Preference Alignment, which leverages learned prototypes as semantic anchors to construct preference pairs, optimizing a domain-compliant rewriting policy without human annotations. At inference time, DAMPER integrates a sampling-based Exponential Mechanism to provide rigorous span-level Differential Privacy (DP) guarantees. Extensive experiments demonstrate that DAMPER significantly outperforms existing baselines, achieving a superior privacy-utility trade-off. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_10145 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Mask-Free Privacy Extraction and Rewriting: A Domain-Aware Approach via Prototype Learning Li, Xiaodong Wang, Yuhua Yu, Qingchen Qin, Zixuan Sun, Yifan Zhang, Qinnan Zhang, Hainan Zheng, Zhiming Cryptography and Security Client-side privacy rewriting is crucial for deploying LLMs in privacy-sensitive domains. However, existing approaches struggle to balance privacy and utility. Full-text methods often distort context, while span-level approaches rely on impractical manual masks or brittle static dictionaries. Attempts to automate localization via prompt-based LLMs prove unreliable, as they suffer from unstable instruction following that leads to privacy leakage and excessive context scrubbing. To address these limitations, we propose DAMPER (Domain-Aware Mask-free Privacy Extraction and Rewriting). DAMPER operationalizes latent privacy semantics into compact Domain Privacy Prototypes via contrastive learning, enabling precise, autonomous span localization. Furthermore, we introduce a Prototype-Guided Preference Alignment, which leverages learned prototypes as semantic anchors to construct preference pairs, optimizing a domain-compliant rewriting policy without human annotations. At inference time, DAMPER integrates a sampling-based Exponential Mechanism to provide rigorous span-level Differential Privacy (DP) guarantees. Extensive experiments demonstrate that DAMPER significantly outperforms existing baselines, achieving a superior privacy-utility trade-off. |
| title | Mask-Free Privacy Extraction and Rewriting: A Domain-Aware Approach via Prototype Learning |
| topic | Cryptography and Security |
| url | https://arxiv.org/abs/2604.10145 |