Ziqiao Wang (汪子乔)

Assistant Professor School of Computer Science and Technology Tongji University Office: Room 405-6, Zhixin Building Address: 4800 Cao'an Highway, Shanghai 201804, China Email: ziqiaowang[AT]tongji[DOT]edu[DOT]cn [CV][Google Scholar][X][BlueSky][Linkedin] Grant me the serenity to accept the things I cannot change, the courage to change the things I can, and the wisdom to know the difference. — Reinhold Niebuhr

About Me

I am a tenure-track assistant professor at School of Computer Science and Technology, Tongji University, and I am also affiliated with the Key Laboratory of Embedded System and Service Computing, Ministry of Education, China. In May 2024, I completed my PhD at the University of Ottawa, where I was fortunate to be advised by Prof. Yongyi Mao. During Summer and Fall 2019, I was a research intern at National Research Council Canada (NRC).

Research Interests

I am broadly interested in the theoretical foundations of machine learning, with a particular focus on statistical learning theory and information theory. My primary research area is generalization theory, including both in-distribution and out-of-distribution settings (e.g., domain adaptation), where I often apply information-theoretic analysis tools. More recently, I have also been exploring safety and trustworthiness in large language models (LLMs), including topics such as LLM alignment and watermarking.

Open Positions

I’m actively seeking highly self-motivated graduate students with a solid foundation in mathematics or empirical deep learning skills to join my group. If you are interested in working with me, please do not hesitate to contact me to inquire about available positions.

If you are an undergraduate student at Tongji and are interested in collaborating with me on research projects, please feel free to reach out.

Selected Publications (Full List)

* denotes equal contribution.

• Generalization in Federated Learning: A Conditional Mutual Information Framework
  Ziqiao Wang, Cheng Long, and Yongyi Mao
  ICML 2025

• Generalization Bounds via Conditional \(f\)-Information
  Ziqiao Wang and Yongyi Mao
  NeurIPS 2024

• On \(f\)-Divergence Principled Domain Adaptation: An Improved Framework
  Ziqiao Wang and Yongyi Mao
  NeurIPS 2024

• Two Facets of SDE Under an Information-Theoretic Lens: Generalization of SGD via Training Trajectories and via Terminal States
  Ziqiao Wang and Yongyi Mao
  UAI 2024

• Sample-Conditioned Hypothesis Stability Sharpens Information-Theoretic Generalization Bounds
  Ziqiao Wang and Yongyi Mao
  NeurIPS 2023

• Tighter Information-Theoretic Generalization Bounds from Supersamples
  Ziqiao Wang and Yongyi Mao
  ICML 2023 (Oral)

• Information-Theoretic Analysis of Unsupervised Domain Adaptation
  Ziqiao Wang and Yongyi Mao
  ICLR 2023

• Over-Training with Mixup May Hurt Generalization
  Zixuan Liu*, Ziqiao Wang*, Hongyu Guo, and Yongyi Mao
  ICLR 2023

• On the Generalization of Models Trained with SGD: Information-Theoretic Bounds and Implications
  Ziqiao Wang and Yongyi Mao
  ICLR 2022

Professional Service

• Conference Reviewer: NeurIPS, ICML, ICLR, AISTATS, UAI, AAAI, IJCAI, ISIT
  

• Journal Reviewer: IEEE Transactions on Information Theory (TIT), Transactions on Machine Learning Research (TMLR), IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), IEEE Journal on Selected Areas in Information Theory (JSAIT), IEEE Transactions on Neural Networks and Learning Systems (TNNLS), Information and Inference: A Journal of the IMA (IMAIAI)
  

• Area Chair: ICLR 2026

• Organizing Committee Member: IEEE North American School of Information Theory (NASIT) 2024, The 2025 China Embodied AI Conference (CEAI)

Miscellaneous

My Erdős number is 4 (Wang \(\to\) Bu \(\to\) Polyanskiy \(\to\) Alon \(\to\) Erdős).