THE-X: Privacy-Preserving Transformer Inference with Homomorphic Encryption

Citations Over TimeTop 10% of 2022 papers

Abstract

As more and more pre-trained language models adopt on-cloud deployment, the privacy issues grow quickly, mainly for the exposure of plain-text user data (e.g., search history, medical record, bank account). Privacy-preserving inference of transformer models is on the demand of cloud service users. To protect privacy, it is an attractive choice to compute only with ciphertext in homomorphic encryption (HE). However, enabling pre-trained models inference on ciphertext data is difficult due to the complex computations in transformer blocks, which are not supported by current HE tools yet. In this work, we introduce THE-X, an approximation approach for transformers, which enables privacy-preserving inference of pre-trained models developed by popular frameworks. THE-X proposes a workflow to deal with complex computation in transformer networks, including all the non-polynomial functions like GELU, softmax, and Layer-Norm. Experiments reveal our proposed THE-X can enable transformer inference on encrypted data for different downstream tasks, all with negligible performance drop but enjoying the theory-guaranteed privacy-preserving advantage.

Related Papers

• → An Efficient Fully Homomorphic Encryption Scheme for Private Information Retrieval in the Cloud(2019)4 cited
• → Performance analysis of private information retrieval scheme based on homomorphic encryption(2015)2 cited
• → HETAL: Efficient Privacy-preserving Transfer Learning with Homomorphic Encryption(2024)8 cited
• → A Pervasive, Efficient and Private Future: Realizing Privacy-Preserving Machine Learning Through Hybrid Homomorphic Encryption(2024)1 cited
• → RETRACTED: Secured learning through electronic health records using Hybrid Fully Homomorphic Encryption(2024)