Welcome to my personal homepage! I am Chuwei Wang, a Ph.D. student from the Department of Computing and Mathematical Sciences (CMS), Caltech. I am fortunate to be advised by Professor Anima Anandkumar and Andrew Stuart. Prior to joining Caltech in 2023, I obtained my B.S. degree in computational mathematics from the School of Mathematical Sciences (SMS), Peking University, where I was advised by Professor Liwei Wang and Di He.

My research goal is to develop practical and powerful machine learning tools for scientific problems through theory, algorithm design, and high performance computing. I am also fascinated with infinite-dimensional objects.

Specifically, I am currently interested in:

• Operator learning.
• Application of deep learning in physical sciences.
• Closure modeling (a.k.a. coarse-grained modeling, reduce order modeling, etc).
• Sampling and optimal transport.

I love working with collaborators from various scientific backgrounds. If you are interested in my research, please feel free to contact me!

News

• Nov. 2024 Awarded as Top Reviewer of NeurIPS 2024 (8.57%, 1300 of 15160 reviewers).
• Aug. 2024 Give a talk on data-driven closure models at IAIFI summer workshop.
• Aug. 2024 Our new work on closure modeling is released on arxiv. We reveal the fundamental shortcomings of existing learning-based closure models and propose a new learning framework, achieving both tremendous speedup and higher accuracy.
• May 2024 We extend Forward Laplacian to accomodate general high-order differential operators. The paper appears at ICLR 2024 workshop on AI4DifferentialEquations.
• Feb. 2024 Our Forward Laplacian paper is accepted by Nature Machine Intelligence. We achieves more than 20x speedup on ground-state energy computing (QMC) for qunatum systems with \(\sim10^2\) electrons.

Selected Recent Works

Beyond Closure Models: Learning Chaotic-Systems via Physics-Informed Neural Operators

Chuwei Wang, Julius Berner, Zongyi Li, Di Zhou, Jiayun Wang, Jane Bae, Anima Anandkumar
Preprint

DOF: Accelerating High-order Differential Operators with Forward Propagation

Ruichen Li*, Chuwei Wang*, Haotian Ye*, Di He, Liwei Wang
ICLR 2024 workshop on AI4DifferentialEquations

A computational framework for neural network-based variational Monte Carlo with Forward Laplacian

Ruichen Li*, Haotian Ye*, Du Jiang, Xuelan Wen, Chuwei Wang, Zhe Li, Xiang Li, Di He, Ji Chen, Weiluo Ren, Liwei Wang
Nature Machine Intelligence 2024

Is L^2 Physics-Informed Loss Always Suitable for Training Physics-Informed Neural Network?

Chuwei Wang*, Shanda Li*, Di He, Liwei Wang
NeurIPS 2022

Services

• Reviewer of JMLR, Nature Machine Intelligence (NMI), Neurips, ICLR.
• Organizing Committee member of Keller Colloquium.

Selected Awards

• Nov. 2024 Top Reviewer of NeurIPS 2024 (8.57%, 1300 of 15160 reviewers).
• Jun. 2023 Peking University Excellent Graduate.
• Aug. 2022 Bronze Medal (ranked No. 6 in total), S.T. Yau College Student Mathematics Competition, analysis and partial differential equations individual.
• May 2021 Elite undergraduate training program of Applied Mathematics and Statistics.
• Dec. 2020 First Prize, Chinese National College Student Physics Competition
• Dec. 2020 First Prize (ranked No. 14 in total), Chinese National College Student Mathematics Competition