Trigger Inversion

Trigger inversion is a technique used to detect and remove backdoors from machine learning models, particularly deep neural networks, that have been maliciously modified to exhibit unintended behavior when presented with specific "trigger" inputs. Current research focuses on improving the efficiency and effectiveness of trigger inversion algorithms, often employing optimization techniques like gradient-based methods and joint optimization across multiple modalities (e.g., image and text). Successfully identifying and neutralizing these backdoors is crucial for ensuring the security and reliability of AI systems across various applications, ranging from code analysis to visual question answering, where compromised models could have significant real-world consequences.

Papers

August 8, 2024

Eliminating Backdoors in Neural Code Models via Trigger Inversion
Weisong Sun, Yuchen Chen, Chunrong Fang, Yebo Feng, Yuan Xiao, An Guo, Quanjun Zhang, Yang Liu, Baowen Xu, Zhenyu Chen
Backdoor Attack Backdoor Defense Backdoor Removal Neural Code Model Insecure Code Trigger Inversion

April 19, 2024

LSP Framework: A Compensatory Model for Defeating Trigger Reverse Engineering via Label Smoothing Poisoning
Beichen Li, Yuanfang Guo, Heqi Peng, Yangxi Li, Yunhong Wang
Backdoor Attack Backdoor Defense Backdoor Trigger Label Poisoning Attack Trigger Inversion Compensatory Strategy

August 7, 2023

TIJO: Trigger Inversion with Joint Optimization for Defending Multimodal Backdoored Models
Indranil Sur, Karan Sikka, Matthew Walmer, Kaushik Koneripalli, Anirban Roy, Xiao Lin, Ajay Divakaran, Susmit Jha
Backdoor Defense Backdoor Trigger Joint Optimization Multimodal Backdoor Trigger Inversion

April 5, 2023

UNICORN: A Unified Backdoor Trigger Inversion Framework
Zhenting Wang, Kai Mei, Juan Zhai, Shiqing Ma
Backdoor Trigger Many Unicorn Backdoor Model Trigger Inversion

January 29, 2023

Gradient Shaping: Enhancing Backdoor Attack Against Reverse Engineering
Rui Zhu, Di Tang, Siyuan Tang, Guanhong Tao, Shiqing Ma, Xiaofeng Wang, Haixu Tang
Backdoor Attack Backdoor Defense Reverse Engineering Gradient Manipulation Trigger Inversion

May 9, 2022

Model-Contrastive Learning for Backdoor Defense
Zhihao Yue, Jun Xia, Zhiwei Ling, Ming Hu, Ting Wang, Xian Wei, Mingsong Chen
Contrastive Learning Backdoor Attack Backdoor Defense Trigger Inversion

February 11, 2022

Constrained Optimization with Dynamic Bound-scaling for Effective NLPBackdoor Defense
Guangyu Shen, Yingqi Liu, Guanhong Tao, Qiuling Xu, Zhuo Zhang, Shengwei An, Shiqing Ma, Xiangyu Zhang
Backdoor Attack Softmax Function Constrained Optimization Bound Propagation Trigger Inversion