Shadow in the Cache: Unveiling and Mitigating Privacy Risks of KV-cache in LLM Inference
缓存中的阴影:揭示和缓解LLM推理中KV缓存的隐私风险
机构 * State Key Laboratory of Blockchain and Data Security(区块链与数据安全国家重点实验室) ; Zhejiang University(浙江大学) ; Huawei Technology(华为技术) ; Hangzhou High-Tech Zone (Binjiang) Institute of Blockchain and Data Security(杭州高新技术区(滨江)区块链与数据安全研究院)
专题命中 隐私与版权 :trustworthy(abstract);分类 cs.CL、cs.AI
AI总结 本文揭示了KV缓存中存在严重的隐私泄露问题,设计了三种攻击方法并提出KV-Cloak防御机制,有效缓解了隐私风险且不影响模型精度。
Comments This paper is accepted by Network and Distributed System Security Symposium (NDSS) 2026. Code: https://github.com/SiO-2/kvcloak
Journal ref Published in the Proceedings of the 33rd Network and Distributed System Security Symposium (NDSS 2026)