Biography
I am now an Assistant Professor in School of Computer Science and Engineering of Nanjing University of Science and Technology, where I am a member of IMAG. I received my doctoral degree from Shanghai Jiao Tong University under the supervision of Dr. Quanshi Zhang in 2023, and received my B.E. degree from Xiamen University in 2018. Morever, I have served as the reviewer for top-tier conferences and journals, including TPAMI, NeurIPS, ICLR, ICML, CVPR, ICCV, AAAI and TMLR.
Research Interests
My research interests include various topics in deep learning, especially the explainablity of deep learning.
Work related to these topics has publised in top-tier conferences and journals, including TPAMI, NeurIPS, ICML, CVPR, ICCV, and AAAI.
Looking for Students and Partnership
I am looking for highly motivated students (Ph.D, master, undergraduate students) to work together on machine learning and Explainable AI, especially theoretical/experimental explanation of the inference logic/representation capacity of DNNs, and reliable guidance for improving the performance. Please feel free to send me your CV (xcheng8@njust.edu.cn), if you have interest.
I am actively looking for partnership so that we can together make the impossible possible :)
招收硕士生、本科实习生，欢迎对人工智能充满兴趣的学生联系我，课题组内具有浓厚的学术氛围，我将为每一位学生供系统而深入的科研指导、丰厚的科研补贴与奖励、充足的计算资源等，优秀者可以支持海外学术交流与工作推荐， 请将简历发送至邮箱xcheng8@njust.edu.cn。
Selected Publications
A Unified Interpretation of Training-Time Out-of-Distribution Detection in ICCV 2025.
Xu Cheng, Xin Jiang, Zechao Li
Layerwise Change of Knowledge in Neural Networks in ICML 2024.
Xu Cheng#, Lei Cheng#, Zhaoran Peng, Yang Xu, Tian Han, Quanshi Zhang
Clarifying the Behavior and the Difficulty of Adversarial Training in AAAI 2024.
Xu Cheng, Hao Zhang, Yue Xin, Wen Shen, Quanshi Zhang
Quantifying the knowledge in a DNN to explain knowledge distillation in TPAMI 2022.
Quanshi Zhang#, Xu Cheng#, Yilan Chen, Zhefan Rao
Explaining knowledge distillation by quantifying the knowledge in CVPR 2020.
Xu Cheng, Zhefan Rao, Yilan Chen, Quanshi Zhang
Interpretable Rotation-Equivariant Multiary-Valued Network for Attribute Obfuscation in TPAMI 2025.
Quanshi Zhang, Hao Zhang, Yiting Chen, Qihan Ren, Jie Ren, Xu Cheng, Liyao Xiang
Towards the Difficulty for a Deep Neural Network to Learn Concepts of Different Complexities in NeurIPS 2023.
Dongrui Liu#, Huiqi Deng#, Xu Cheng, Qihan Ren, Kangrui Wang, Quanshi Zhang
A Unified Game-theoretic Interpretation of Adversarial Robustness in NeurIPS 2021.
Jie Ren#, Die Zhang#, Yisen Wang#, Lu Chen, Zhanpeng Zhou, Yiting Chen, Xu Cheng, Xin Wang, Meng Zhou, Jie Shi, Quanshi Zhang.
Building Interpretable Interaction Trees for Deep NLP Models in AAAI 2021.
Die Zhang, Huilin Zhou, Hao Zhang, Xiaoyi Bao, Da Huo, Ruizhao Chen, Xu Cheng, Mengyue Wu, Quanshi Zhang
ArXiv Papers
Formulating and Analyzing How a DNN Encodes Visual Concepts from the Interaction Perspective in arXiv:2106.10938.
Xu Cheng, Chuntung Chu, Yi Zheng, Jie Ren, Jinpeng Zhang, Hao Zhang, Quanshi Zhang
Do Game-Theoretic Interactions Really Reflect Prototypical Concepts in DNNs? in arXiv:2108.02646.
Xu Cheng, Xin Wang, Haotian Xue, Zhengyang Liang, Quanshi Zhang