Trigger Design

Trigger design in machine learning focuses on creating subtle patterns (triggers) that manipulate model behavior, often to launch backdoor attacks. Current research investigates trigger design across various model architectures, including deep neural networks, large language models, diffusion models, and even the neural architecture search process itself, exploring both visible and invisible, static and dynamic trigger types. This research is crucial for understanding and mitigating vulnerabilities in AI systems, impacting the security and trustworthiness of applications ranging from image recognition and natural language processing to federated learning and generative AI. The development of robust detection and defense mechanisms against these attacks is a key area of ongoing investigation.

Papers

May 9, 2024

Poisoning-based Backdoor Attacks for Arbitrary Target Label with Positive Triggers
Binxiao Huang, Jason Chun Lok, Chang Liu, Ngai Wong
Backdoor Attack Backdoor Defense Adversarial Method Backdoor Poisoning Attack Trigger Design Target Class Trigger Warning

May 5, 2024

Trojans in Large Language Models of Code: A Critical Review through a Trigger-Based Taxonomy
Aftab Hussain, Md Rafiqul Islam Rabin, Toufique Ahmed, Bowen Xu, Premkumar Devanbu, Mohammad Amin Alipour
Large Language Model Real World Code Comprehensive Taxonomy Critical Review Code Model Trojan Attack Trigger Design

February 5, 2024

DisDet: Exploring Detectability of Backdoor Attack on Diffusion Models
Yang Sui, Huy Phan, Jinqi Xiao, Tianfang Zhang, Zijie Tang, Cong Shi, Yan Wang, Yingying Chen, Bo Yuan
Diffusion Model Backdoor Attack Information Theoretic Detectability Trigger Design

November 28, 2023

Rethinking Backdoor Attacks on Dataset Distillation: A Kernel Method Perspective
Ming-Yu Chung, Sheng-Yen Chou, Chia-Mu Yu, Pin-Yu Chen, Sy-Yen Kuo, Tsung-Yi Ho
Backdoor Attack Dataset Distillation Kernel Method Effective Backdoor Attack Trigger Design Resilient Backdoor

November 26, 2023

BadCLIP: Trigger-Aware Prompt Learning for Backdoor Attacks on CLIP
Jiawang Bai, Kuofeng Gao, Shaobo Min, Shu-Tao Xia, Zhifeng Li, Wei Liu
Backdoor Attack Prompt Learning Single CLIP Pre Trained CLIP Trigger Design

October 1, 2023

GhostEncoder: Stealthy Backdoor Attacks with Dynamic Triggers to Pre-trained Encoders in Self-supervised Learning
Qiannan Wang, Changchun Yin, Zhe Liu, Liming Fang, Run Wang, Chenhao Lin
Self Supervised Learning Backdoor Attack Encoder Side Stealthy Backdoor Attack Backdoor Sample Invisible Backdoor Attack Trigger Design

August 31, 2023

FTA: Stealthy and Adaptive Backdoor Attack with Flexible Triggers on Federated Learning
Yanqi Qiao, Dazhuang Liu, Congwen Chen, Rui Wang, Kaitai Liang
Digital STEALTH Metric Trigger Design Robust Backdoor Attack Imperceptible Trigger

June 1, 2023

Versatile Backdoor Attack with Visible, Semantic, Sample-Specific, and Compatible Triggers
Ruotong Wang, Hongrui Chen, Zihao Zhu, Li Liu, Baoyuan Wu
Deep Neural Network Backdoor Attack Natural Trigger Trigger Design Trigger Point

March 25, 2023

Backdoor Attacks with Input-unique Triggers in NLP
Xukun Zhou, Jiwei Li, Tianwei Zhang, Lingjuan Lyu, Muqiao Yang, Jun He
Natural Language Processing Backdoor Attack Backdoor Trigger Backdoor Injection Trigger Design Backdoor Policy

November 20, 2022

Invisible Backdoor Attack with Dynamic Triggers against Person Re-identification
Wenli Sun, Xinyang Jiang, Shuguang Dou, Dongsheng Li, Duoqian Miao, Cheng Deng, Cairong Zhao
Adversarial Attack Backdoor Attack Person Re Identification Invisible Backdoor Attack Trigger Design

November 16, 2022

Deep Intention-Aware Network for Click-Through Rate Prediction
Yaxian Xia, Yi Cao, Sihao Hu, Tong Liu, Lingling Lu
Click Through Rate Prediction Intent Prediction User Intent CTR Prediction Trigger Design

October 21, 2022

Neural Architectural Backdoors
Ren Pang, Changjiang Li, Zhaohan Xi, Shouling Ji, Ting Wang
Neural Architecture Search Attack Vector Hidden Backdoor Neural Backdoor Trigger Design

Trigger Design

Papers

Poisoning-based Backdoor Attacks for Arbitrary Target Label with Positive Triggers

Trojans in Large Language Models of Code: A Critical Review through a Trigger-Based Taxonomy

DisDet: Exploring Detectability of Backdoor Attack on Diffusion Models

Rethinking Backdoor Attacks on Dataset Distillation: A Kernel Method Perspective

BadCLIP: Trigger-Aware Prompt Learning for Backdoor Attacks on CLIP

GhostEncoder: Stealthy Backdoor Attacks with Dynamic Triggers to Pre-trained Encoders in Self-supervised Learning

FTA: Stealthy and Adaptive Backdoor Attack with Flexible Triggers on Federated Learning

Versatile Backdoor Attack with Visible, Semantic, Sample-Specific, and Compatible Triggers

Backdoor Attacks with Input-unique Triggers in NLP

Invisible Backdoor Attack with Dynamic Triggers against Person Re-identification

Deep Intention-Aware Network for Click-Through Rate Prediction

Neural Architectural Backdoors