EncFormer reduces online MPC communication by 1.4x-30.4x and end-to-end latency by 1.3x-9.8x versus prior hybrid FHE-MPC systems for private GPT- and BERT-style inference while preserving accuracy.
Membership inference attacks against machine learning models,
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CR 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
PrivFedTalk presents a federated conditional latent diffusion system with LoRA identity adapters, ISFA aggregation, and TDC regularization that enables privacy-preserving personalized talking-head generation across distributed clients.
citing papers explorer
-
EncFormer: Secure and Efficient Transformer Inference over Encrypted Data
EncFormer reduces online MPC communication by 1.4x-30.4x and end-to-end latency by 1.3x-9.8x versus prior hybrid FHE-MPC systems for private GPT- and BERT-style inference while preserving accuracy.
-
PrivFedTalk: Privacy-Aware Federated Diffusion with Identity-Stable Adapters for Personalized Talking-Head Generation
PrivFedTalk presents a federated conditional latent diffusion system with LoRA identity adapters, ISFA aggregation, and TDC regularization that enables privacy-preserving personalized talking-head generation across distributed clients.