About me

Iā€™m a third year Ph.D. student in Applied Mathematics at Stony Brook University supervised by Professor Yi Liu.

Previously, I obtained a Bachelor of Science degree from Stony Brook University in both applied mathematicsšŸ“™ and pure mathematicsšŸ“•.

My research focuses on Geometric Deep Learning, Generative Models, Neural Operators, and AI for Science in general.

The best way to contact me is through email: wenhan(dot)gao(at symbol)stonybrook(dot)edu šŸ“«

Education

  • Computer Science and Information Security (3.97/4.0); Queensborough Community College, Aug. 2018 - May 2019
  • B.S. Mathematics; Applied Mathematics & Statistics (3.97/4.0); Stony Brook University, Aug. 2019 - Dec. 2021
  • Ph.D. Operations Research (4.0/4.0); Stony Brook University, Aug. 2022 - TBD

News

  • January 2025: First author on a paper accepted by ICLR 2025 on discretization mismatch errors in neural operators.
  • December 2024: Accepted an offer from Visa Research to work as an intern staff research scientist over the summer.
  • November 2024: Passed the preliminary exam in AMS. Slides
  • October 2024: Received the NeurIPS Scholar Award, see you in Vancouver.
  • September 2024: Co-first author on a paper, in collaboration with FSU, accepted by NeurIPS on symmetries for molecular GNN active learning!
  • September 2024: First author on a paper accepted by TMLR on symmetries in neural operators!
  • August 2024: Finished teaching AMS 326, Numerical Analysis. Congrats to all my students on a job well done! Syllabus;Lecture Notes
  • August 2024: Mentored undergraduate student Xiang Liu (freshman in CS/AMS), who successfully completed a summer research project on neural operators for climate change through the SUNY SOAR program. Congrats!
  • March 2024: Awarded the Excellence in Student Teaching Award from the AMS department at Stony Brook for Fall 2023!
  • December 2023: Finished teaching the graduate course AMS 595, Fundamentals of Computing. Congrats to all my students on a job well done! Syllabus;Lecture Notes
  • August 2023: Passed Ph.D. qualifying exam in AMS!
  • February 2023: First author on a paper on active learning-based sampling for high-dimensional PDEs published in the Journal of Computational Physics.

First or Co-first Author Publications

* indicates equal contributors.

Discretization-invariance? On the Discretization Mismatch Errors in Neural Operators

W.Gao, R.Xu, Y.Deng, Y.Liu.

The Thirteenth International Conference on Learning Representations (ICLR), 2025

Paper | Code | BibTeX

Empowering Active Learning for 3D Molecular Graphs with Geometric Graph Isomorphism

R.Subedi*, L.Wei, W.Gao*, S.Chakraborty, Y.Liu.

The Thirty-eighth Annual Conference on Neural Information Processing Systems (NeurIPS), 2024

Paper | Code | BibTeX

Coordinate Transform Fourier Neural Operators for Symmetries in Physical Modelings

W.Gao, R.Xu, H.Wang, Y.Liu.

Transactions on Machine Learning Research (TMLR), 2024

Paper | Code | BibTeX

First or Co-first Author Preprints

  • One paper that reveals neural operators can learn hidden physics from data; In Submission to ICML 25; W.Gao*, J.Luo*, R.Xu, F.Wan, X.Liu, Y.Liu
  • One paper that proposes a principled explanation method for 3D GNNs (explainable AI); In Submission to ICML 25; J.Qu*, W.Gao*, J.Zhang, X.Liu, H.Wei, H.Ling, Y.Liu
  • One paper addressing the misalignment between optimized masks and actual explanatory subgraphs in 3D GNNs; In Submission; X.Liu*, W.Gao*, Y.Liu
  • One paper on designing enhanced kernels for localized effects and interactions in FNO; In Submission; W.Gao, J.Luo, R.Xu, Y.Liu