Encrypted Large Model Inference: The Equivariant Encryption Paradigm
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912216966496256 |
|---|---|
| author | Buban, James Zhang, Hongyang Angione, Claudio Yang, Harry Farhan, Ahmad Sultanov, Seyfal Du, Michael Ma, Xuran Wang, Zihao Zhao, Yue Owlia, Arria Johnston, Fielding Colangelo, Patrick |
| author_facet | Buban, James Zhang, Hongyang Angione, Claudio Yang, Harry Farhan, Ahmad Sultanov, Seyfal Du, Michael Ma, Xuran Wang, Zihao Zhao, Yue Owlia, Arria Johnston, Fielding Colangelo, Patrick |
| contents | Large scale deep learning model, such as modern language models and diffusion architectures, have revolutionized applications ranging from natural language processing to computer vision. However, their deployment in distributed or decentralized environments raises significant privacy concerns, as sensitive data may be exposed during inference. Traditional techniques like secure multi-party computation, homomorphic encryption, and differential privacy offer partial remedies but often incur substantial computational overhead, latency penalties, or limited compatibility with non-linear network operations. In this work, we introduce Equivariant Encryption (EE), a novel paradigm designed to enable secure, "blind" inference on encrypted data with near zero performance overhead. Unlike fully homomorphic approaches that encrypt the entire computational graph, EE selectively obfuscates critical internal representations within neural network layers while preserving the exact functionality of both linear and a prescribed set of non-linear operations. This targeted encryption ensures that raw inputs, intermediate activations, and outputs remain confidential, even when processed on untrusted infrastructure. We detail the theoretical foundations of EE, compare its performance and integration complexity against conventional privacy preserving techniques, and demonstrate its applicability across a range of architectures, from convolutional networks to large language models. Furthermore, our work provides a comprehensive threat analysis, outlining potential attack vectors and baseline strategies, and benchmarks EE against standard inference pipelines in decentralized settings. The results confirm that EE maintains high fidelity and throughput, effectively bridging the gap between robust data confidentiality and the stringent efficiency requirements of modern, large scale model inference. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_01013 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Encrypted Large Model Inference: The Equivariant Encryption Paradigm Buban, James Zhang, Hongyang Angione, Claudio Yang, Harry Farhan, Ahmad Sultanov, Seyfal Du, Michael Ma, Xuran Wang, Zihao Zhao, Yue Owlia, Arria Johnston, Fielding Colangelo, Patrick Cryptography and Security Artificial Intelligence Large scale deep learning model, such as modern language models and diffusion architectures, have revolutionized applications ranging from natural language processing to computer vision. However, their deployment in distributed or decentralized environments raises significant privacy concerns, as sensitive data may be exposed during inference. Traditional techniques like secure multi-party computation, homomorphic encryption, and differential privacy offer partial remedies but often incur substantial computational overhead, latency penalties, or limited compatibility with non-linear network operations. In this work, we introduce Equivariant Encryption (EE), a novel paradigm designed to enable secure, "blind" inference on encrypted data with near zero performance overhead. Unlike fully homomorphic approaches that encrypt the entire computational graph, EE selectively obfuscates critical internal representations within neural network layers while preserving the exact functionality of both linear and a prescribed set of non-linear operations. This targeted encryption ensures that raw inputs, intermediate activations, and outputs remain confidential, even when processed on untrusted infrastructure. We detail the theoretical foundations of EE, compare its performance and integration complexity against conventional privacy preserving techniques, and demonstrate its applicability across a range of architectures, from convolutional networks to large language models. Furthermore, our work provides a comprehensive threat analysis, outlining potential attack vectors and baseline strategies, and benchmarks EE against standard inference pipelines in decentralized settings. The results confirm that EE maintains high fidelity and throughput, effectively bridging the gap between robust data confidentiality and the stringent efficiency requirements of modern, large scale model inference. |
| title | Encrypted Large Model Inference: The Equivariant Encryption Paradigm |
| topic | Cryptography and Security Artificial Intelligence |
| url | https://arxiv.org/abs/2502.01013 |