许志钦Zhi-Qin John Xu

Zhi-Qin John Xu is a professor at the Institute of Natural Sciences/School of Mathematical Sciences, Shanghai Jiao Tong University. Zhi-Qin graduated from Zhiyuan College of Shanghai Jiao Tong University in 2012. In 2016, he graduated from Shanghai Jiao Tong University with a doctor's degree in applied mathematics. From 2016 to 2019, he was a postdoctoral fellow at NYU ABU Dhabi and the Courant Institute. From 2019 to 2025, he was an associate professor at Shanghai Jiao Tong University. For language model, he identifies the complexity of model is critical to the memorization and reason capbility of a language model. In deep learning theory, he and collaborators discovered frequency principle, parameter condensation and embedding principles in deep learning, and developed multi-scale neural networks; In AI for Science, mainly solving PDE and ODE, such as combustion, he and collaborators developed deep learning based mechanism reduction (DeePMR) and deep learning based surrogate model for accelerating the simulation of chemical kinetics (DeePODE). He published papers as the first author or corresponding author at TPAMI, JMLR, NeurIPS，ICML, ICLR, AAAI， SIMODS, CiCP, CSIAM Trans. Appl. Math., JCP, Combustion and Flame, Eur. J. Neurosci. etc. Currently, he is the Managing Editor of Journal of Machine Learning.

Understanding Deep Learning Principles from Phenomena
• Large Language Models: Investigating the memory and reasoning capabilities of large language models.
• Deep Learning Theory: Frequency principle, parameter condensation and embedding principles, and multiscale neural networks.
• AI for Science: Solving PDEs and ODEs, particularly multiscale problems such as the Fokker-Planck problem and combustion problems.

Founding Executive Editor of the Journal of Machine Learning (JML). Editors-in-Chief: Weinan E, Jianfeng Lu. It publishes high-quality research papers across all areas of machine learning, covering algorithms, theory, and their applications in artificial intelligence, science, and engineering, with equal emphasis on both theory and applications. Fully open access and no fees are charged. Sponsored by the China Society for Industrial and Applied Mathematics (CSIAM), the Center for Machine Learning Research at Peking University, and the Beijing Institute of Scientific Intelligence, and published by Global Science Press.