Publications

(* denotes equal contribution)

2025

  1. S&P
    BAIT: Large Language Model Backdoor Scanning by Inverting Attack Target
    Guangyu Shen, Siyuan Cheng, Zhuo Zhang, Guanhong Tao, Kaiyuan Zhang, Hanxi Guo, Lu Yan, Xiaolong Jin, Shengwei An, Shiqing Ma, and Xiangyu Zhang
    In Proceedings of the 46th IEEE Symposium on Security and Privacy, San Francisco, CA, USA, 2025

2024

  1. NeurIPS
    BiScope: AI-generated Text Detection by Checking Memorization of Preceding Tokens
    Hanxi Guo, Siyuan Cheng, Xiaolong Jin, Zhuo Zhang, Kaiyuan Zhang, Guanhong Tao, Guangyu Shen, and Xiangyu Zhang
    In Proceedings of Thirty-Eighth Conference on Neural Information Processing Systems, Vancouver, Canada, 2024
  2. ACSAC
    Exploring Inherent Backdoors in Deep Learning Models
    Guanhong Tao, Siyuan Cheng, Zhenting Wang, Shiqing Ma, Shengwei An, Yingqi Liu, Guangyu Shen, Zhuo Zhang, Yunshu Mao, and Xiangyu Zhang
    In Proceedings of the Annual Computer Security Applications Conference, Hawaii, USA, 2024
  3. ECCV
    UNIT: Backdoor Mitigation via Automated Neural Distribution Tightening
    Siyuan Cheng, Guangyu Shen, Kaiyuan Zhang, Guanhong Tao, Shengwei An, Hanxi Guo, Shiqing Ma, and Xiangyu Zhang
    In Proceedings of the 18th European Conference on Computer Vision, Milan, Italy, 2024
  4. USENIX Security
    Rethinking the Invisible Protection against Unauthorized Image Usage in Stable Diffusion
    Shengwei An, Lu Yan, Siyuan Cheng, Guangyu Shen, Kaiyuan Zhang, Qiuling Xu, Guanhong Tao, and Xiangyu Zhang
    In Proceedings of the 33rd USENIX Security Symposium, Philadelphia, PA, USA, 2024
  5. S&P
    Distribution Preserving Backdoor Attack in Self-supervised Learning
    Guanhong Tao*, Zhenting Wang*, Shiwei Feng, Guangyu Shen, Shiqing Ma, and Xiangyu Zhang
    In Proceedings of the 45th IEEE Symposium on Security and Privacy, San Francisco, CA, USA, 2024
  6. S&P
    ODSCAN: Backdoor Scanning for Object Detection Models
    Siyuan Cheng, Guangyu Shen, Guanhong Tao, Kaiyuan Zhang, Zhuo Zhang, Shengwei An, Xiangzhe Xu, Yingqi Liu, Shiqing Ma, and Xiangyu Zhang
    In Proceedings of the 45th IEEE Symposium on Security and Privacy, San Francisco, CA, USA, 2024
  7. S&P
    On Large Language Models’ Resilience to Coercive Interrogation
    Zhuo Zhang, Guangyu Shen, Guanhong Tao, Siyuan Cheng, and Xiangyu Zhang
    In Proceedings of the 45th IEEE Symposium on Security and Privacy, San Francisco, CA, USA, 2024
  8. S&P
    Exploring the Orthogonality and Linearity of Backdoor Attacks
    Kaiyuan Zhang, Siyuan Cheng, Guangyu Shen, Guanhong Tao, Shengwei An, Anuran Makur, Shiqing Ma, and Xiangyu Zhang
    In Proceedings of the 45th IEEE Symposium on Security and Privacy, San Francisco, CA, USA, 2024
  9. NDSS
    Gradient Shaping: Enhancing Backdoor Attack Against Reverse Engineering
    Rui Zhu, Di Tang, Siyuan Tang, Zihao Wang, Guanhong Tao, Shiqing Ma, Xiaofeng Wang, and Haixu Tang
    In Proceedings of the 31st Network and Distributed System Security Symposium, San Diego, CA, USA, 2024
  10. CVPR
    LOTUS: Evasive and Resilient Backdoor Attacks through Sub-Partitioning
    Siyuan Cheng, Guanhong Tao, Yingqi Liu, Guangyu Shen, Shengwei An, Shiwei Feng, Xiangzhe Xu, Kaiyuan Zhang, Shiqing Ma, and Xiangyu Zhang
    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle WA, USA, 2024
  11. AAAI
    Elijah: Eliminating Backdoors Injected in Diffusion Models via Distribution Shift
    Shengwei An, Sheng-Yen Chou, Kaiyuan Zhang, Qiuling Xu, Guanhong Tao, Guangyu Shen, Siyuan Cheng, Shiqing Ma, Pin-Yu Chen, Tsung-Yi Ho, and  others
    In Proceedings of the 38th AAAI Conference on Artificial Intelligence, Vancouver, Canada, 2024
  12. ICLR
    Fusion is Not Enough: Single Modal Attack on Fusion Models for 3D Object Detection
    Zhiyuan Cheng, Hongjun Choi, Shiwei Feng, James Chenhao Liang, Guanhong Tao, Dongfang Liu, Michael Zuzak, and Xiangyu Zhang
    In Proceedings of the Twelfth International Conference on Learning Representations, Vienna, Austria, 2024
  13. EACL
    Threat Behavior Textual Search by Attention Graph Isomorphism
    Chanwoo Bae, Guanhong Tao, Zhuo Zhang, and Xiangyu Zhang
    In Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers), St. Julians, Malta, 2024

2023

  1. USENIX Security
    Hard-label Black-box Universal Adversarial Patch Attack
    Guanhong Tao, Shengwei An, Siyuan Cheng, Guangyu Shen, and Xiangyu Zhang
    In Proceedings of the 32nd USENIX Security Symposium, Anaheim, CA, USA, 2023
  2. USENIX Security
    PELICAN: Exploiting Backdoors of Naturally Trained Deep Learning Models In Binary Code Analysis
    Zhuo Zhang, Guanhong Tao, Guangyu Shen, Shengwei An, Qiuling Xu, Yingqi Liu, Yapeng Ye, Yaoxuan Wu, and Xiangyu Zhang
    In Proceedings of the 32nd USENIX Security Symposium, Anaheim, CA, USA, 2023
  3. ACL
    Backdooring Neural Code Search
    Weisong Sun*, Yuchen Chen*, Guanhong Tao*, Chunrong Fang, Xiangyu Zhang, Quanjun Zhang, and Bin Luo
    In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics, Toronto, Canada, 2023
  4. NDSS
    BEAGLE: Forensics of Deep Learning Backdoor Attack for Better Defense
    Siyuan Cheng, Guanhong Tao, Yingqi Liu, Shengwei An, Xiangzhe Xu, Shiwei Feng, Guangyu Shen, Kaiyuan Zhang, Qiuling Xu, Shiqing Ma, and  others
    In Proceedings of the 30th Network and Distributed System Security Symposium, San Diego, CA, USA, 2023
  5. ICLR
    FLIP: A Provable Defense Framework for Backdoor Mitigation in Federated Learning
    Kaiyuan Zhang, Guanhong Tao, Qiuling Xu, Siyuan Cheng, Shengwei An, Yingqi Liu, Shiwei Feng, Guangyu Shen, Pin-Yu Chen, Shiqing Ma, and  others
    In Proceedings of the Eleventh International Conference on Learning Representations, Kigali, Rwanda, 2023
  6. ICLR
    Adversarial Training of Self-supervised Monocular Depth Estimation against Physical-World Attacks
    Zhiyuan Cheng, James Chenhao Liang, Guanhong Tao, Dongfang Liu, and Xiangyu Zhang
    In Proceedings of the Eleventh International Conference on Learning Representations, Kigali, Rwanda, 2023
  7. NeurIPS
    BIRD: Generalizable Backdoor Detection and Removal for Deep Reinforcement Learning
    Xuan Chen, Wenbo Guo, Guanhong Tao, Xiangyu Zhang, and Dawn Song
    In Proceedings of Thirty-seventh Conference on Neural Information Processing Systems, New Orleans, LA, USA, 2023
  8. NeurIPS
    Django: Detecting Trojans in Object Detection Models via Gaussian Focus Calibration
    Guangyu Shen, Siyuan Cheng, Guanhong Tao, Kaiyuan Zhang, Yingqi Liu, Shengwei An, Shiqing Ma, and Xiangyu Zhang
    In Proceedings of Thirty-seventh Conference on Neural Information Processing Systems, New Orleans, LA, USA, 2023
  9. NeurIPS
    ParaFuzz: An Interpretability-Driven Technique for Detecting Poisoned Samples in NLP
    Lu Yan, Zhuo Zhang, Guanhong Tao, Kaiyuan Zhang, Xuan Chen, Guangyu Shen, and Xiangyu Zhang
    In Proceedings of Thirty-seventh Conference on Neural Information Processing Systems, New Orleans, LA, USA, 2023
  10. CVPR
    MEDIC: Remove Model Backdoors via Importance Driven Cloning
    Qiuling Xu, Guanhong Tao, Jean Honorio, Yingqi Liu, Shengwei An, Guangyu Shen, Siyuan Cheng, and Xiangyu Zhang
    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Vancouver, Canada, 2023
  11. CVPR
    Detecting Backdoors in Pre-trained Encoders
    Shiwei Feng, Guanhong Tao, Siyuan Cheng, Guangyu Shen, Xiangzhe Xu, Yingqi Liu, Kaiyuan Zhang, Shiqing Ma, and Xiangyu Zhang
    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Vancouver, Canada, 2023
  12. S&P
    ImU: Physical Impersonating Attack for Face Recognition System with Natural Style Changes
    Shengwei An, Yuan Yao, Qiuling Xu, Shiqing Ma, Guanhong Tao, Siyuan Cheng, Kaiyuan Zhang, Yingqi Liu, Guangyu Shen, Ian Kelk, and  others
    In Proceedings of the 44th IEEE Symposium on Security and Privacy, San Francisco, CA, USA, 2023
  13. FSE
    PEM: Representing Binary Program Semantics for Similarity Analysis via a Probabilistic Execution Model
    Xiangzhe Xu, Zhou Xuan, Shiwei Feng, Siyuan Cheng, Yapeng Ye, Qingkai Shi, Guanhong Tao, Le Yu, Zhuo Zhang, and Xiangyu Zhang
    In Proceedings of the 31st ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering, San Francisco, California, USA, 2023
  14. ISSTA
    Improving Binary Code Similarity Transformer Models by Semantics-Driven Instruction Deemphasis
    Xiangzhe Xu, Shiwei Feng, Yapeng Ye, Guangyu Shen, Zian Su, Siyuan Cheng, Guanhong Tao, Qingkai Shi, Zhuo Zhang, and Xiangyu Zhang
    In Proceedings of the 32nd ACM SIGSOFT International Symposium on Software Testing and Analysis, Seattle, Washington, USA, 2023

2022

  1. S&P
    Model Orthogonalization: Class Distance Hardening in Neural Networks for Better Security
    Guanhong Tao, Yingqi Liu, Guangyu Shen, Qiuling Xu, Shengwei An, Zhuo Zhang, and Xiangyu Zhang
    In Proceedings of the 43rd IEEE Symposium on Security and Privacy, San Francisco, CA, USA, 2022
  2. S&P
    PICCOLO: Exposing Complex Backdoors in NLP Transformer Models
    Yingqi Liu, Guangyu Shen, Guanhong Tao, Shengwei An, Shiqing Ma, and Xiangyu Zhang
    In Proceedings of the 43rd IEEE Symposium on Security and Privacy, San Francisco, CA, USA, 2022
  3. FSE
    RULER: Discriminative and Iterative Adversarial Training for Deep Neural Network Fairness
    Guanhong Tao*, Weisong Sun*, Tingxu Han*, Chunrong Fang, and Xiangyu Zhang
    In Proceedings of the 30th ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering, Singapore, 2022
  4. CVPR
    Better Trigger Inversion Optimization in Backdoor Scanning
    Guanhong Tao, Guangyu Shen, Yingqi Liu, Shengwei An, Qiuling Xu, Shiqing Ma, Pan Li, and Xiangyu Zhang
    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, LA, USA, 2022
  5. CVPR
    Bounded Adversarial Attack on Deep Content Features
    Qiuling Xu, Guanhong Tao, and Xiangyu Zhang
    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, LA, USA, 2022
  6. CVPR
    Complex Backdoor Detection by Symmetric Feature Differencing
    Yingqi Liu, Guangyu Shen, Guanhong Tao, Zhenting Wang, Shiqing Ma, and Xiangyu Zhang
    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, LA, USA, 2022
  7. NDSS
    MIRROR: Model Inversion for Deep Learning Network with High Fidelity
    Shengwei An, Guanhong Tao, Qiuling Xu, Yingqi Liu, Guangyu Shen, Yuan Yao, Jingwei Xu, and Xiangyu Zhang
    In Proceedings of the 29th Network and Distributed System Security Symposium, San Diego, CA, USA, 2022
  8. ICML
    Constrained Optimization with Dynamic Bound-scaling for Effective NLP Backdoor Defense
    Guangyu Shen, Yingqi Liu, Guanhong Tao, Qiuling Xu, Zhuo Zhang, Shengwei An, Shiqing Ma, and Xiangyu Zhang
    In Proceedings of the 39th International Conference on Machine Learning, Baltimore, MD, USA, 2022
  9. ECCV
    Physical Attack on Monocular Depth Estimation in Autonomous Driving with Optimal Adversarial Patches
    Zhiyuan Cheng, James Liang, Hongjun Choi, Guanhong Tao, Zhiwen Cao, Dongfang Liu, and Xiangyu Zhang
    In Proceedings of the 2022 European Conference on Computer Vision, Tel Aviv, Israel, 2022
  10. CAIN
    Checkpointing and Deterministic Training for Deep Learning
    Xiangzhe Xu, Hongyu Liu, Guanhong Tao, Zhou Xuan, and Xiangyu Zhang
    In Proceedings of the 1st International Conference on AI Engineering: Software Engineering for AI, Pittsburgh, PA, USA, 2022
  11. ICSE
    Code Search based on Context-aware Code Translation
    Weisong Sun, Chunrong Fang, Yuchen Chen, Guanhong Tao, Tingxu Han, and Quanjun Zhang
    In Proceedings of the 44th International Conference on Software Engineering, Pittsburgh, PA, USA, 2022

2021

  1. ICLR Workshop
    FIRM: Detecting Adversarial Audios by Recursive Filters with Randomization
    Guanhong Tao, Xiaowei Chen, Yunhan Jia, Zhenyu Zhong, Shiqing Ma, and Xiangyu Zhang
    In ICLR 2021 Workshop on Security and Safety in Machine Learning Systems, Virtual Event, 2021
  2. AAAI
    Towards Feature Space Adversarial Attack by Style Perturbation
    Qiuling Xu, Guanhong Tao, Siyuan Cheng, and Xiangyu Zhang
    In Proceedings of the 35th AAAI Conference on Artificial Intelligence, Virtual Event, 2021
  3. ICML
    Backdoor Scanning for Deep Neural Networks through K-Arm Optimization
    Guangyu Shen, Yingqi Liu, Guanhong Tao, Shengwei An, Qiuling Xu, Siyuan Cheng, Shiqing Ma, and Xiangyu Zhang
    In Proceedings of Thirty-eighth International Conference on Machine Learning, Virtual Event, 2021
  4. S&P
    StochFuzz: Sound and Cost-effective Fuzzing of Stripped Binaries by Incremental and Stochastic Rewriting
    Zhuo Zhang, Wei You, Guanhong Tao, Yousra Aafer, Xuwei Liu, and Xiangyu Zhang
    In Proceedings of the 42nd IEEE Symposium on Security and Privacy, Virtual Event, 2021
  5. S&P
    OSPREY: Recovery of Variable and Data Structure via Probabilistic Analysis for Stripped Binary
    Zhuo Zhang, Yapeng Ye, Wei You, Guanhong Tao, Wen-chuan Lee, Yonghwi Kwon, Yousra Aafer, and Xiangyu Zhang
    In Proceedings of the 42nd IEEE Symposium on Security and Privacy, Virtual Event, 2021
  6. NDSS
    ALchemist: Fusing Application and Audit Logs for Precise Attack Provenance without Instrumentation
    Le Yu, Shiqing Ma, Zhuo Zhang, Guanhong Tao, Xiangyu Zhang, Dongyan Xu, Vincent E Urias, Han Wei Lin, Gabriela Ciocarlie, Vinod Yegneswaran, and Ashish Gehani
    In Proceedings of the 28th Network and Distributed System Security Symposium, Virtual Event, 2021

2020

  1. ICSE
    TRADER: Trace Divergence Analysis and Embedding Regulation for Debugging Recurrent Neural Networks
    Guanhong Tao, Shiqing Ma, Yingqi Liu, Qiuling Xu, and Xiangyu Zhang
    In Proceedings of the ACM/IEEE 42nd International Conference on Software Engineering, Seoul, South Korea, 2020
  2. ICSE
    CPC: Automatically Classifying and Propagating Natural Language Comments via Program Analysis
    Juan Zhai, Xiangzhe Xu, Yu Shi, Guanhong Tao, Minxue Pan, Shiqing Ma, Lei Xu, Weifeng Zhang, Lin Tan, and Xiangyu Zhang
    In Proceedings of the ACM/IEEE 42nd International Conference on Software Engineering, Seoul, South Korea, 2020
  3. FSE
    Correlations Between Deep Neural Network Model Coverage Criteria and Model Quality
    Shenao Yan, Guanhong Tao, Xuwei Liu, Juan Zhai, Shiqing Ma, Lei Xu, and Xiangyu Zhang
    In Proceedings of the 28th ACM Joint Meeting on European Software Engineering Conference and Symposium on the Foundations of Software Engineering, Virtual Event, 2020

2019

  1. CCS
    ABS: Scanning Neural Networks for Back-doors by Artificial Brain Stimulation
    Yingqi Liu, Wen-Chuan Lee, Guanhong Tao, Shiqing Ma, Yousra Aafer, and Xiangyu Zhang
    In Proceedings of the 26th ACM Conference on Computer and Communications Security, London, United Kingdom, 2019
  2. NDSS
    NIC: Detecting Adversarial Samples with Neural Network Invariant Checking
    Shiqing Ma, Yingqi Liu, Guanhong Tao, Wen-Chuan Lee, and Xiangyu Zhang
    In Proceedings of the 26th Network and Distributed System Security Symposium, San Diego, CA, USA, 2019
  3. OOPSLA
    BDA: Practical Dependence Analysis for Binary Executables by Unbiased Whole-program Path Sampling and Per-path Abstract Interpretation
    Zhuo Zhang, Wei You, Guanhong Tao, Guannan Wei, Yonghwi Kwon, and Xiangyu Zhang
    Proceedings of the ACM on Programming Languages, Athens, Greece, 2019

2018

  1. NeurIPS
    Attacks Meet Interpretability: Attribute-steered Detection of Adversarial Samples
    Guanhong Tao, Shiqing Ma, Yingqi Liu, and Xiangyu Zhang
    In Proceedings of Thirty-second Conference on Neural Information Processing Systems, Montréal, Canada, 2018
  2. CCS
    Precise Android API Protection Mapping Derivation and Reasoning
    Yousra Aafer, Guanhong Tao, Jianjun Huang, Xiangyu Zhang, and Ninghui Li
    In Proceedings of the 25th ACM Conference on Computer and Communications Security, Toronto, Canada, 2018
  3. IEEE Trans. Reliab.
    MalPat: Mining Patterns of Malicious and Benign Android Apps via Permission-Related APIs
    Guanhong Tao, Zibin Zheng, Ziying Guo, and Michael R. Lyu
    IEEE Transactions on Reliability, 2018