Zhengdao Chen

PhD in Mathematics, Courant Institute of Mathematical Sciences, New York University
(Google Scholar) (GitHub)

Publications and Preprints

  A. Deep Learning Theory

Neural Hilbert Ladders: Multi-Layer Neural Networks in Function Space
Journal of Machine Learning Research, 25(109), 1−65
Zhengdao Chen

Multi-Layer Neural Networks as Trainable Ladders of Hilbert Spaces
40th International Conference on Machine Learning (ICML 2023)
Zhengdao Chen

A Functional-Space Mean-Field Theory of Partially-Trained Three-Layer Neural Networks
arxiv:2210.16286
Zhengdao Chen, Eric Vanden-Eijnden, Joan Bruna

On Feature Learning in Neural Networks with Global Convergence Guarantees
10th International Conference on Learning Representations (ICLR 2022)
Zhengdao Chen, Eric Vanden-Eijnden, Joan Bruna

A Dynamical Central Limit Theorem for Shallow Neural Networks
34th Conference on Neural Information Processing Systems (NeurIPS 2020)
Zhengdao Chen, Grant M. Rotskoff, Joan Bruna, Eric Vanden-Eijnden

  B. Graph Neural Networks

A Non-Asymptotic Analysis of Oversmoothing in Graph Neural Networks
11th International Conference on Learning Representations (ICLR 2023)
Xinyi Wu, Zhengdao Chen, William Wang, Ali Jadbabaie

Learning the Relevant Substructures for Tasks on Graph Data
46th IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2021)
Lei Chen, Zhengdao Chen, Joan Bruna

On Graph Neural Networks versus Graph-Augmented MLPs
9th International Conference on Learning Representations (ICLR 2021)
Lei Chen*, Zhengdao Chen*, Joan Bruna (* indicates joint authorship)

Can Graph Neural Networks Count Substructures?
34th Conference on Neural Information Processing Systems (NeurIPS 2020)
Zhengdao Chen, Lei Chen, Soledad Villar, Joan Bruna

On the Equivalence between Graph Isomorphism Testing and Function Approximation with GNNs  (code)
33rd Conference on Neural Information Processing Systems (NeurIPS 2019)
Zhengdao Chen, Soledad Villar, Lei Chen, Joan Bruna

Supervised Community Detection with Line Graph Neural Networks  (code)
7th International Conference on Learning Representations (ICLR 2019)
Zhengdao Chen, Lisha Li, Joan Bruna

  C. Learning and Simulating Dynamics

Symplectic Recurrent Neural Networks  (code)
8th International Conference on Learning Representations (ICLR 2020, spotlight)
Zhengdao Chen, Jianyu Zhang, Martin Arjovsky, Léon Bottou

Structure-preserving numerical integrators for Hodgkin-Huxley-type systems
SIAM Journal on Scientific Computing, 42(1), B273-B298
Zhengdao Chen, Baranidharan Raman, Ari Stern

  D. Older Topics

Statistical learning and spelling: Older prephonological spellers produce more wordlike spellings than younger prephonological spellers
Child Development, 89(4), e431-e443
Rebecca Treiman, Brett Kessler, Kelly Boland, Hayley Clocksin, Zhengdao Chen

Multimodal surface matching with higher-order smoothness constraints
NeuroImage, 167, 453-465
Emma C Robinson, Kara Garcia, Matthew F Glasser, Zhengdao Chen, Timothy S Coalson, Antonios Makropoulos, Jelena Bozek, Robert Wright, Andreas Schuh, Matthew Webster, Jana Hutter, Anthony Price, Lucilio Cordero Grande, Emer Hughes, Nora Tusor, Philip V Bayly, David C Van Essen, Stephen M Smith, A David Edwards, Joseph Hajnal, Mark Jenkinson, Ben Glocker, Daniel Rueckert

Engaging and disengaging recurrent inhibition coincides with sensing and unsensing of a sensory stimulus
Nature Communications, 8, 15413
Debajit Saha, Wensheng Sun, Chao Li, Srinath Nizampatnam, William Padovano, Zhengdao Chen, Alex Chen, Ege Altan, Ray Lo, Dennis L Barbour, Baranidharan Raman