TY - GEN
T1 - PPQformer
T2 - 2025 2nd International Conference on Generative Artificial Intelligence and Information Security, GAIIS 2025
AU - Dong, Benchang
AU - Chen, Zhili
N1 - Publisher Copyright:
© 2025 Copyright held by the owner/author(s).
PY - 2025/6/3
Y1 - 2025/6/3
N2 - In the context of the popularity of Transformer model services, the issue of privacy protection has gradually gained attention. However, the primary issues with existing private inference solutions lie in memory constraints and suboptimal computational performance. In this work, we introduce PPQformer, a framework that leverages Replicated Secret Sharing (RSS) and quantization techniques to enable efficient and secure inference of Transformer models in a Multi-Party Computation (MPC) setting. By integrating quantization into the MPC protocols, PPQformer significantly reduces memory footprint and enhances computational performance while preserving data and model privacy. Experimental results demonstrate the effectiveness of our approach, achieving competitive accuracies with reduced communication cost and inference time compared to existing methods.
AB - In the context of the popularity of Transformer model services, the issue of privacy protection has gradually gained attention. However, the primary issues with existing private inference solutions lie in memory constraints and suboptimal computational performance. In this work, we introduce PPQformer, a framework that leverages Replicated Secret Sharing (RSS) and quantization techniques to enable efficient and secure inference of Transformer models in a Multi-Party Computation (MPC) setting. By integrating quantization into the MPC protocols, PPQformer significantly reduces memory footprint and enhances computational performance while preserving data and model privacy. Experimental results demonstrate the effectiveness of our approach, achieving competitive accuracies with reduced communication cost and inference time compared to existing methods.
KW - Multi-Party Computation
KW - Privacy-preserving model inference
KW - Replicated Secret Sharing
KW - Transformer models quantization
UR - https://www.scopus.com/pages/publications/105015572425
U2 - 10.1145/3728725.3728743
DO - 10.1145/3728725.3728743
M3 - 会议稿件
AN - SCOPUS:105015572425
T3 - Proceedings of the 2025 2nd International Conference on Generative Artificial Intelligence and Information Security, GAIIS 2025
SP - 115
EP - 119
BT - Proceedings of the 2025 2nd International Conference on Generative Artificial Intelligence and Information Security, GAIIS 2025
PB - Association for Computing Machinery, Inc
Y2 - 21 February 2025 through 23 February 2025
ER -