Jian Cui

Lecturer, School of Cyber Science and Technology, Beihang University

1 paper at NeurIPS 2025

Homepage· OpenReview· Semantic Scholar· Google Scholar

Poster Session 2

1 paper

Wednesday, December 3, 2025 · 4:30 PM → 7:30 PM

Exhibit Hall C,D,E

LoRO: Real-Time on-Device Secure Inference for LLMs via TEE-Based Low Rank Obfuscation

#1404 · Gaojian Xiong, Yu Sun, Jianhua Liu, Jian Cui, Jianwei Liu

We analyze and compromise existing TEE-based secure inference methods by our Model Stealing attacks with Prior, and propose a novel secure inference method named LoRO, which is efficient, secure and accurate.