arxiv.org - The-X Privacy-Preserving Transformer Inference With Homomorphic Encryption

The-X: Privacy-Preserving Transformer Inference With Homomorphic Encryption - arxiv.org ![rw-book-cover|200x400](https://readwise-assets.s3.amazonaws.com/static/images/article1.be68295a7e40.png) ## Metadata - Author: **arxiv.org** - Full Title: The-X: Privacy-Preserving Transformer Inference With Homomorphic Encryption - Category: #articles - Tags: #ai - URL: https://arxiv.org/abs/2206.00216 ## Highlights - THE-X proposes a workflow to deal with complex computation in transformer networks, including all the non-polynomial functions like GELU, softmax, and LayerNorm. 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.