Private Transformer Inference in MLaaS: A Survey
It tackles privacy risks for users and providers in MLaaS deployments, but is incremental as a survey and taxonomy.
This paper surveys Private Transformer Inference (PTI) in Machine Learning as a Service (MLaaS) to address privacy concerns from centralized data processing, reviewing cryptographic techniques like secure multi-party computation and homomorphic encryption to enable private inference while preserving data and model privacy.
Transformer models have revolutionized AI, powering applications like content generation and sentiment analysis. However, their deployment in Machine Learning as a Service (MLaaS) raises significant privacy concerns, primarily due to the centralized processing of sensitive user data. Private Transformer Inference (PTI) offers a solution by utilizing cryptographic techniques such as secure multi-party computation and homomorphic encryption, enabling inference while preserving both user data and model privacy. This paper reviews recent PTI advancements, highlighting state-of-the-art solutions and challenges. We also introduce a structured taxonomy and evaluation framework for PTI, focusing on balancing resource efficiency with privacy and bridging the gap between high-performance inference and data privacy.