SecureTLM: private inference for transformer-based large model with MPC

DOI10.1016/J.INS.2024.120429MaRDI QIDQ6544553zbMATH OpenFDO

Authors Yuntian Chen, Xianjia Meng, Zhiying Shi, Zhiyuan Ning, Jingzhi Lin

Publication date 27 May 2024

Published in Information Sciences (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1016/J.INS.2024.120429

zbMATH Keywords

hybrid secret sharingprivate inferencesecure multiparty computationtransformer

Mathematics Subject Classification ID

Privacy of data (68P27) Modes of computation (nondeterministic, parallel, interactive, probabilistic, etc.) (68Q10) Cryptography (94A60) Authentication, digital signatures and secret sharing (94A62)

Recommendations

• Privacy preserving multi-party computation delegation for deep learning in cloud computing
• Efficient and privacy-preserving tree-based inference via additive homomorphic encryption
• Multiparty homomorphic machine learning with data security and model preservation
• SecureBiNN: 3-party secure computation for binarized neural network inference

Cites work

• A novel privacy-preserving graph convolutional network via secure matrix multiplication
• Efficient and privacy-preserving tree-based inference via additive homomorphic encryption
• How to simulate it -- a tutorial on the simulation proof technique
• scientific article; zbMATH DE number 176564 (Why is no real title available?)
• Privacy-enhanced and non-interactive linear regression with dropout-resilience

This page was built for publication: SecureTLM: private inference for transformer-based large model with MPC

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6544553)