1 citations · 1 across the 6 of their papers we have counts for
5 papers · 1 filter
TrajGuard: Streaming Hidden-state Trajectory Detection for Decoding-time Jailbreak Defense
Cheng Liu, Xiaolei Liu, Xingyu Li +2
Existing jailbreak defense paradigms primarily rely on static detection of prompts, outputs, or internal states, often neglecting the dynamic evolution of risk during decoding. Thi…
LoopLLM: Transferable Energy-Latency Attacks in LLMs via Repetitive Generation
Xingyu Li, Xiaolei Liu, Cheng Liu +4
As large language models (LLMs) scale, their inference incurs substantial computational resources, exposing them to energy-latency attacks, where crafted prompts induce high energy…
Segmented Private Data Aggregation in the Multi-message Shuffle Model
Shaowei Wang, Hongqiao Chen, Sufen Zeng +8
The shuffle model of differential privacy (DP) offers compelling privacy-utility trade-offs in decentralized settings (e.g., internet of things, mobile edge networks). Particularly…
INK: Inheritable Natural Backdoor Attack Against Model Distillation
Xiaolei Liu, Ming Yi, Kangyi Ding +4
Deep learning models are vulnerable to backdoor attacks, where attackers inject malicious behavior through data poisoning and later exploit triggers to manipulate deployed models.…
Fine-grained Private Knowledge Distillation
Yuntong Li, Shaowei Wang, Yingying Wang +4
Knowledge distillation has emerged as a scalable and effective way for privacy-preserving machine learning. One remaining drawback is that it consumes privacy in a model-level (i.e…