Xukun Luan
2 indexed papers
Publications per year
Top categories
Frequent co-authors
Research Timeline
This paper proposes PA-HDP, a framework for privacy-preserving retrieval-augmented generation using prompt-aware dynamic hierarchical differential privacy.
This paper introduces Code-Poisoning Property Inference Attack (CPPIA), the first code-level Property Inference Attack (PIA) for leaking privacy from ML models trained on private data, overcoming limitations of existing works.
Papers
Code-Poisoning Property Inference Attacks
Xukun Luan, Yuhui Gong, Gang Zhang, Zixuan Huang +3 more
This paper introduces Code-Poisoning Property Inference Attack (CPPIA), the first code-level Property Inference Attack (PIA) for leaking privacy from ML models trained on private data, overcoming limi…