Tai Liang

PASTA proposes a novel, twofold stealthy backdoor attack that enables high-success-rate backdoor activation across arbitrary patches in Vision Transformers by leveraging the Trigger Radiating Effect (TRE).

DETOUR proposes a practical backdoor attack against object detection models by using semantic triggers that are robust to variations in size, location, and field of view (FoV), overcoming limitations of existing fixed-trigger attacks.

The paper introduces PassNet, a large-scale ecosystem for generating compiler passes using LLMs, demonstrating that LLMs can significantly accelerate graph compilation for long-tail workloads, suggesting that consistency is the primary bottleneck.