Publications
Papers, chapters, preprints, and my doctoral thesis, ordered from newest to oldest.
Years indicate publication or first preprint appearance.
-
Propagation of Chaos in Contextual Flow Maps
Accepted · NeurIPS 2026 [Conference] [arXiv]
-
Scaling Limits of Long-Context Transformers
Accepted · NeurIPS 2026 [Conference] [arXiv]
-
Critical Attention Scaling in Long-Context Transformers
ICLR 2026 [Conference] [arXiv] [OpenReview] [Poster]
-
Correcting Autodifferentiation in Neural ODE Training
SIAM Journal on Applied Mathematics 86(5), 2360–2378 · 2026 [Journal] [arXiv]
-
Quantitative Clustering in Mean-Field Transformer Models
Accepted · Courant Journal of Pure and Applied Mathematics; arXiv:2504.14697 · 2025 [arXiv]
-
Residual Connections Provably Mitigate Oversmoothing in Graph Neural Networks
Accepted · Journal of Machine Learning Research; arXiv:2501.00762 · 2025 [arXiv]
-
Accelerating Optimization over the Space of Probability Measures
Journal of Machine Learning Research 26(31), 1–40 · 2025 [Journal] [arXiv]
-
Multiscale Numerical Methods for Elliptic and Wave-Type PDEs and Their Inverse Problems
Doctoral thesis · University of Wisconsin–Madison, 2024 [Thesis]
-
Bayesian Sampling Using Interacting Particles
Book chapter · Active Particles, Volume 4: Theory, Models, Applications, 175–215 · 2024 [Book chapter] [arXiv]
-
A Good Score Does Not Lead to a Good Generative Model
-
A Reduced Order Schwarz Method for Nonlinear Multiscale Elliptic Equations Based on Two-Layer Neural Networks
Journal of Computational Mathematics 42(2), 570–596 · 2024 [Journal] [arXiv] [GitHub]
-
On Optimal Bases for Multiscale PDEs and Bayesian Homogenization
Preprint · arXiv:2305.12303 · 2023 [arXiv]
-
Learning Harmonic Molecular Representations on Riemannian Manifold
ICLR 2023, poster [Conference] [arXiv] [OpenReview] [GitHub] [Poster]
-
High-Frequency Limit of the Inverse Scattering Problem: Asymptotic Convergence from Inverse Helmholtz to Inverse Liouville
SIAM Journal on Imaging Sciences 16(1), 111–143 · 2023 [Journal] [arXiv] [GitHub] [Poster]
-
Overparameterization of Deep ResNet: Zero Loss and Mean-Field Analysis
Journal of Machine Learning Research 23(48), 1–65 · 2022 [Journal] [arXiv]
-
Low-Rank Approximation for Multiscale PDEs
Notices of the American Mathematical Society 69(6) · 2022 [Journal] [arXiv]
-
Manifold Learning and Nonlinear Homogenization
Multiscale Modeling & Simulation 20(3), 1093–1126 · 2022 [Journal] [arXiv] [GitHub]
-
On the Global Convergence of Gradient Descent for Multi-Layer ResNets in the Mean-Field Regime
Preprint · arXiv:2110.02926 · 2021 [arXiv]
-
Semiclassical Limit of an Inverse Problem for the Schrödinger Equation
Research in the Mathematical Sciences 8, article 39 · 2021 [Journal] [arXiv]
-
Classical Limit for the Varying-Mass Schrödinger Equation with Random Inhomogeneities
Journal of Computational Physics 438, article 110365 · 2021 [Journal] [arXiv]
-
State-Specific Projection of COVID-19 Infection in the United States and Evaluation of Three Major Control Measures
Scientific Reports 10, article 22429 · 2020 [Journal] [medRxiv] [GitHub] [Poster]
-
How Implementation of Entropy in Driving Structural Ordering of Nanoparticles Relates to Assembly Kinetics: Insight into Reaction-Induced Interfacial Assembly of Janus Nanoparticles
Langmuir 34(32), 9477–9488 · 2018 [Journal]