Publication List

2025 [Preprint] Xudong Pan, Jiarun Dai, Yihe Fan, Min Yang. Frontier AI systems have surpassed the self-replicating red line. Arxiv. [Link] [Preprint] Changyue Jiang, Xudong Pan, Geng Hong, Chenfu Bao, Min Yang. RAG-Thief: Scalable Extraction of Private Data from Retrieval-Augmented Generation Applications with Agent-based Attacks. Arxiv. [Link] 2024 [TPAMI] Xudong Pan, Mi Zhang, Yifan Yan, Shengyao Zhang, Min Yang. Matryoshka: Exploiting the Over-Parametrization of Deep Learning Models for Covert Data Transmission. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), (IF=24.314), 2024. [Link] [S&P'24] Huming Qiu, Junjie Sun, Mi Zhang, Xudong Pan, Min Yang. BELT: Old-School Backdoor Attacks can Evade the State-of-the-Art Defense with Backdoor Exclusivity Lifting, the 2024 IEEE Symposium on Security and Privacy (S&P). 2024. [Link] [TDSC'24] Chenghui Shi, Shouling Ji, Xudong Pan, Xuhong Zhang, Mi Zhang, Min Yang, Jun Zhou, Jianwei Yin, Ting Wang. Towards Practical Backdoor Attacks on Federated Learning Systems. IEEE Transactions on Dependable and Secure Computing (TDSC), 2024. [Link] [CCS'24] Neural Dehydration: Effective Erasure of Black-box Watermarks from DNNs with Limited Data. Yifan Lu, Wenxuan Li, Mi Zhang, Xudong Pan, Min Yang. The 31th ACM Conference on Computer and Communications Security (CCS), 2024. [Link] 2023 [KDD'23] Xudong Pan, Mi Zhang, Yifan Yan, Yining Wang, Min Yang. Cracking White-box DNN Watermarks via Invariant Neuron Transforms, the 29th SIGKDD Conference on Knowledge Discovery and Data Mining (KDD), P1783–1794, 2023. [Link] [USENIX Security'23] Yifan Yan (*), Xudong Pan (*), Mi Zhang, Min Yang. Rethinking White-Box Watermarks on Deep Learning Models under Neural Structural Obfuscation, the 32nd USENIX Security Symposium (USENIX Security, *equal contribution), P2347-2364, 2023. [Link] [USENIX Security'23] Qifan Xiao (*), Xudong Pan (*), Yifan Lu, Mi Zhang, Min Yang. Exorcising “Wraith”: Protecting LiDAR-based Object Detector in Automated Driving System from Appearing Attacks, the 32nd USENIX Security Symposium (USENIX Security, *equal contribution), P2939-2956, 2023. [Link] [ACL'23] Shengyao Zhang, Xudong Pan, Mi Zhang, Min Yang. SlowBERT: Slow-down Attacks on Input-adaptive Multi-exit BERT. Findings of the Association for Computational Linguistics (ACL), P9992–10007, 2023. [Link] [ICASSP'23] Xudong Pan, Mi Zhang, Duocai Wu. RØROS: Building a Responsive Online Recommender System via Meta-Gradients Updating, 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2023. [Link] [WWW'23] Xiaoyu You, Beina Sheng, Daizong Ding, Mi Zhang, Xudong Pan, Min Yang, Fuli Feng. MaSS: Model-agnostic, Semantic and Stealthy Data Poisoning Attack on Knowledge Graph Embedding, the 2023 Web Conference (WWW), P2000-2010, 2023. [Link] [WWW'23] Xiaoyu You, Chi Lee, Daizong Ding, Mi Zhang, Fuli Feng, Xudong Pan, Min Yang. Anti-FakeU: Defending Shilling Attacks on Graph Neural Network based Recommender Model, the 2023 Web Conference (WWW), P938-948, 2023. [Link] 2022 [NeurIPS'22] Xudong Pan, Shengyao Zhang, Mi Zhang, Yifan Yan, Min Yang. House of Cans: Covert Transmission of Internal Datasets via Capacity-Aware Neuron Steganography, the 36th Annual Conference on Neural Information Processing Systems (NeurIPS), 2022. [Link]