Edge-placed QR-inspired structured patches reduce infrared CLIP classifier accuracy from 98.67% to 0.70% and transfer to black-box captioning and VQA models.
SecBERT: Privacy-preserving pre-training based neural network inference system
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2representative citing papers
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.
citing papers explorer
-
InfraQR: Edge-Placed QR-Inspired Structured Patch Attacks on Infrared Vision-Language Models
Edge-placed QR-inspired structured patches reduce infrared CLIP classifier accuracy from 98.67% to 0.70% and transfer to black-box captioning and VQA models.
-
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.