I am an Assistant Professor in the Department of Applied Mathematics and Statistics and the Department of Computer Science at Colorado School of Mines. I am also part of the Mines Optimization and Deep Learning (MODL) group. Prior to joining Mines, I was an Assistant Adjunct Professor in the Department of Mathematics at UCLA. I received my PhD in applied mathematics from Emory University in 2019, where I worked under the guidance of Lars Ruthotto. My research interests lie in the intersection of applied mathematics and data science. In particular, I am interested in inverse problems, optimization, optimal control, and deep learning.

If you are a self-motivated student interested in working in the aforementioned areas, send me an email.

Recent News

• 12/2023: Our draft Structured World Representations in Maze-Solving Transformers is out and has been accepted by the in NeurIPS Workshop on Unifying Representations in Neural Models (UniReps). Thanks to Michael Ivanitskiy, Alex Spies, Tilman Räuker, and everyone else involved for the collaboration.
• 09/2023: Our preprint A Configurable Library for Generating and Manipulating Maze Datasets is out. Thanks to Michael Ivanitskiy and everyone else involved for the collaboration, and to AI Safety Camp for the support.
• 07/2023: NSF-DMS award received! This grant will help fund our work on learning to optimize. Thanks to my colleagues and Mines for the help and support.
• 06/2023: Our paper, Explainable AI via Learning to Optimize has been accepted by Scientific Reports. Thanks to Howard Heaton for his collaboration.
• 03/2023: It was a great day to have the Mines Optimization and Deep Learning group together.
• 03/2023: Our paper Taming Hyperparameter Tuning in Continuous Normalizing Flows Using the JKO Scheme has been accepted by Scientific Reports. Thanks to Alexander Vidal, Luis Tenorio, Stanley Osher, and Levon Nuberkyan for their collaboration.
• 03/2023: Our paper A Hamilton-Jacobi-based Proximal Operator has been accepted by the Proceedings of the National Academy of Sciences. Thanks to Howard Heaton and Stan Osher for the collaboration.
• 02/2023: Our draft Faster Predict-and-Optimize with Davis-Yin Splitting is out. Thanks to Daniel McKenzie and Howard Heaton for the collaboration.
• 12/2022: Our SIAM News Article on Learning to Optimize is out. Thanks to Daniel McKenzie and Wotao Yin for the collaboration.
• 11/2022: Our paper A Numerical Algorithm for Inverse Problem from Partial Boundary Measurement Arising from Mean Field Game Problem has been accepted by the journal Inverse Problems. Thanks to Yat Tin Chow, Siting Liu, Levon Nurbekyan, and Stan Osher for the collaboration.
• 11/2022: Our paper Global Solutions to Nonconvex Problems via Evolution of Hamilton-Jacobi PDEs has been accepted by the journal Communications on Applied Mathematics and Computation. Thanks to Howard Heaton and Stanley Osher for the collaboration.
• 07/21/2022: It’s been a great summer working at the 2022 Emory Computational Mathematics for Data Science REU/RET on Model Meets Data. Congratulations to Linghai Liu, Lisa Zhou, and Allen Tong (left to right) on a successful REU and poster presentation on implicit deep learning and inverse problems.
• 05/2022: Our draft Explainable AI via Learning to Optimize is out. Thanks to Howard Heaton for the collaboration.
• 04/2022: I am humbled to receive the inaugural 2022 MGB-SIAM Early Career Fellowship.
• 04/2022: Our draft A Numerical Algorithm for Inverse Problem from Partial Boundary Measurement Arising from Mean Field Game Problem is out. Thanks to Yat Tin Chow, Siting Liu, Levon Nurbekyan, and Stan Osher for the collaboration.
• 04/2022: Our draft Adaptive Uncertainty-Weighted ADMM for Distributed Optimization has been accepted by the Journal of Applied and Numerical Optimization. Thanks to Jianping Ye and Caleb Wan for the collaboration.
• 02/2022: Our draft Random Features for High-Dimensional Nonlocal Mean-Field Games has been accepted by the Journal of Computational Physics. Thanks to Sudhanshu Agrawal, Wonjun Lee, and Levon Nurbekyan for the collaboration.

Select Publications

• Heaton H, Wu Fung S. Explainable AI via Learning to Optimize, Scientific Reports, 13 (10103). 2023
• Osher S, Heaton H, Wu Fung S. A Hamilton-Jacobi-based Proximal Operator, Proceedings of the National Academy of Sciences. 2023
• Wu Fung S, Heaton H, Li Q, McKenzie D, Osher S, Yin W. JFB: Jacobian-Free Backpropagation for Implicit Networks, AAAI Conference on Artificial Intelligence, 36(6), 6648-6656. 2022
• Heaton H, McKenzie D, Li Q, Wu Fung S, Osher S, Yin W. Learn to Predict Equilibria via Fixed Point Networks. arXiv:2106.00906. 2021
• Lin AT, Wu Fung S, Li W, Nurbekyan L, Osher S. Alternating the Population and Agent Control via Two Neural Networks to Solve High-Dimensional Stochastic Mean Field Games, Proceedings of the National Academy of Sciences, 118(31). 2021
• Onken D, Wu Fung S, Li X, Ruthotto L. OT-Flow: Fast and Accurate Continuous Normalizing Flows via Optimal Transport, AAAI Conference on Artificial Intelligence, 35(10), 9223-9232. 2021
• Ruthotto L, Osher S, Li W, Nurbekyan L, Wu Fung S. A Machine Learning Framework for Solving High-Dimensional Mean Field Game and Mean Field Control Problems, Proceedings of the National Academy of Sciences, 117(17), 9183-9193. 2020