publications

5555

1. Rigor Over Hype: What Foundation Model Should Mean in Computational Science

   Youngsoo Choi , Siu Wun Cheung , Youngkyu Kim , and 10 more authors

   Computing in Science & Engineering , May 5555

   Abs

   The term foundation model is rapidly entering scientific machine learning, yet its meaning in computational science remains inconsistent. Models are increasingly labeled foundational without evidence of broad generalization across physical systems, geometries, boundary/initial conditions, tasks, or operating regimes. We propose a definition of foundation models tailored to computational science and identify core and desirable characteristics, drawing inspiration from classical foundational methods (FEM/FVM/FDM) that achieve wide reuse through problem-independent structure and principled enforcement of governing equations. We argue that scientific foundation models face unique barriers relative to language and vision models, including field-valued data, mesh and geometry dependence, heterogeneous physics, expensive adaptation, limited benchmarks, and strict requirements for interpretability and physical consistency. To make claims measurable, we introduce a reviewer- and author-friendly checklist for reporting training breadth, generalization axes, adaptation cost, extrapolation protocols, and scientific consistency checks. We also highlight a constructive hybrid path: local learning with global physics enforcement.

2026

1. JCP

   Latent Space Dynamics Identification for Interface Tracking with Application to Shock-Induced Pore Collapse

   Seung Whan Chung, Christopher Miller , Youngsoo Choi , and 3 more authors

   Journal of Computational Physics, May 2026

   Abs

   Capturing sharp, evolving interfaces remains a central challenge in reduced-order modeling, especially when data is limited and the system exhibits localized nonlinearities or discontinuities. We propose LaSDI-IT (Latent Space Dynamics Identification for Interface Tracking), a data-driven framework that combines low-dimensional latent dynamics learning with explicit interface-aware encoding to enable accurate and efficient modeling of physical systems involving moving material boundaries. At the core of LaSDI-IT is a revised autoencoder architecture that jointly reconstructs the physical field and an indicator function representing material regions or phases, allowing the model to track complex interface evolution without requiring detailed physical models or mesh adaptation. The latent dynamics are learned through linear regression in the encoded space and generalized across parameter regimes using Gaussian process interpolation with greedy sampling. We demonstrate LaSDI-IT on the problem of shock-induced pore collapse in high explosives, a process characterized by sharp temperature gradients and dynamically deforming pore geometries. The method achieves relative prediction errors below 9% across the parameter space, accurately recovers key quantities of interest such as pore area and hot spot formation, and matches the performance of dense training with only half the data. This latent dynamics prediction was 106 times faster than the conventional high-fidelity simulation, proving its utility for multi-query applications. These results highlight LaSDI-IT as a general, data-efficient framework for modeling discontinuity-rich systems in computational physics, with potential applications in multiphase flows, fracture mechanics, and phase change problems.

2. Hard-constrained Physics-informed Neural Networks for Interface Problems

   Seung Whan Chung, Stephen Castonguay , Sumanta Roy , and 3 more authors

   May 2026
3. IJNME

   A Scalable Reduced-Order Model for the Steady Navier–Stokes Equations

   Seung Whan Chung, Siu Wun Cheung , Youngsoo Choi , and 7 more authors

   International Journal for Numerical Methods in Engineering, May 2026

   Abs

   ABSTRACT Scaling up new scientific technologies from laboratory to industry often involves demonstrating performance on a larger scale. Computer simulations can accelerate design and predictions in the deployment process, though traditional numerical methods are computationally intractable even for intermediate pilot plant scales. Recently, the component reduced order modeling method has been developed to tackle this challenge by combining projection reduced order modeling and discontinuous Galerkin domain decomposition. However, while many scientific or engineering applications involve nonlinear physics, this method has only been demonstrated for various linear systems. In this work, the component reduced order modeling method is extended to steady Navier–Stokes flow, with application to general nonlinear physics in view. The large-scale, global domain is decomposed into a combination of small-scale unit component. Linear subspaces for flow velocity and pressure are identified via proper orthogonal decomposition over sample snapshots collected from each small-scale unit component. Velocity bases are augmented with a pressure supremizer to satisfy the inf–sup condition for stable pressure prediction. Two different nonlinear reduced order modeling methods are employed and compared for efficient evaluation of nonlinear advection: A third-order tensor projection operator and the empirical quadrature procedure. The proposed method is demonstrated on the flow over arrays of five different unit objects, achieving a 23-fold speedup with less than 4% relative error in domains up to 256 times larger than the unit components. Furthermore, a numerical experiment with the pressure supremizer strongly indicates the need for a supremizer for stable pressure prediction. A comparison between the tensorial approach and the empirical quadrature procedure revealed a slight advantage of the empirical quadrature procedure. The framework is compared with an alternating Schwarz-based reduced-order approach, demonstrating improved efficiency and robustness for the DG-based global solver while retaining flexibility for sub-scale iterative solvers. The method is further extended to a coupled advection–diffusion and Navier–Stokes system, illustrating its applicability to multi-physics problems and its potential for more general, inter-coupled nonlinear systems.

4. Latent Space Element Method

   Seung Whan Chung, Youngsoo Choi , Christopher Miller , and 2 more authors

   May 2026
5. Local Reduced-Order Modeling for Electrostatic Plasmas by Physics-Informed Solution Manifold Decomposition

   Ping-Hsuan Tsai , Seung Whan Chung, Debojyoti Ghosh , and 3 more authors

   Computer Physics Communications, May 2026

   Abs

   Despite advancements in high-performance computing and modern numerical algorithms, computational cost remains prohibitive for multi-query kinetic plasma simulations. In this work, we develop data-driven reduced-order models (ROMs) for collisionless electrostatic plasma dynamics, based on the kinetic Vlasov-Poisson equation. Our ROM approach projects the equation onto a linear subspace defined by the proper orthogonal decomposition (POD) modes. We introduce an efficient tensorial method to update the nonlinear term using a precomputed third-order tensor. We capture multiscale behavior with a minimal number of POD modes by decomposing the solution manifold into multiple time windows and creating temporally local ROMs. We consider two strategies for decomposition: one based on the physical time and the other based on the electric field energy. Applied to the 1D1V Vlasov–Poisson simulations, that is, prescribed E-field, Landau damping, and two-stream instability, we demonstrate that our ROMs accurately capture the total energy of the system both for parametric and time extrapolation cases. The temporally local ROMs are more efficient and accurate than the single ROM. In addition, in the two-stream instability case, we show that the energy-windowing reduced-order model (EW-ROM) is more efficient and accurate than the time-windowing reduced-order model (TW-ROM). With the tensorial approach, EW-ROM solves the equation approximately 90 times faster than Eulerian simulations while maintaining a maximum relative error of 7.5% for the training data and 11% for the testing data.

6. CMAME

   mLaSDI: Multi-stage latent space dynamics identification

   William Anderson , Seung Whan Chung, Robert Stephany , and 1 more author

   Computer Methods in Applied Mechanics and Engineering, May 2026

   Abs

   Accurately solving partial differential equations (PDEs) is essential across many scientific disciplines. However, high-fidelity solvers can be computationally prohibitive, motivating the development of reduced-order models (ROMs). Recently, Latent Space Dynamics Identification (LaSDI) was proposed as a data-driven, non-intrusive ROM framework. LaSDI compresses the training data via an autoencoder and learns user-specified ordinary differential equations (ODEs) governing the latent dynamics, enabling rapid predictions for unseen parameters. While LaSDI has produced effective ROMs for numerous problems, the autoencoder must simultaneously reconstruct the training data and satisfy the imposed latent dynamics. These are often competing objectives which limit accuracy, particularly for complex or high-frequency phenomena. To address this limitation, we propose multi-stage Latent Space Dynamics Identification (mLaSDI). With mLaSDI, we train LaSDI sequentially in stages. After training the initial autoencoder, we train additional decoders which map the latent trajectories to residuals from previous stages. This staged residual learning, combined with periodic activation functions, enables recovery of high-frequency content without sacrificing interpretability of the latent dynamics. We further provide an error decomposition separating autoencoder and latent dynamics contributions, and prove that additional training stages cannot increase the training residual. Numerical experiments on a multiscale oscillating system, unsteady wake flow, and the 1D-1V Vlasov equation demonstrate that mLaSDI achieves significantly lower reconstruction and prediction errors, while requiring less training time and reduced hyperparameter tuning compared to standard LaSDI.

7. CSE

   Rigor Over Hype: What Foundation Model Should Mean in Computational Science

   Youngsoo Choi , Siu Wun Cheung , Youngkyu Kim , and 9 more authors

   May 2026
8. AMC

   Theory and numerics of subspace approximation of eigenvalue problems

   Siu Wun Cheung , Youngsoo Choi , Seung Whan Chung, and 3 more authors

   Applied Mathematics and Computation, May 2026

   Abs

   Large-scale eigenvalue problems arise in various fields of science and engineering and demand computationally efficient solutions. In this study, we investigate the subspace approximation for parametric linear eigenvalue problems, aiming to mitigate the computational burden associated with high-fidelity systems. We provide general error estimates under non-simple eigenvalue conditions, establishing some theoretical foundations for understanding the convergence behavior of subspace approximations. Numerical examples, including problems with one-dimensional to three-dimensional spatial domain and one-dimensional to two-dimensional parameter domain, are presented to demonstrate the efficacy of reduced basis method in handling parametric variations in boundary conditions and coefficient fields to achieve significant computational savings while maintaining high accuracy, making them promising tools for practical applications in large-scale eigenvalue computations.

2025

1. PSST

   Characterization of uncertainties in electron-argon collision cross sections

   Seung Whan Chung, Todd A Oliver , Laxminarayan L Raja , and 1 more author

   Plasma Sources Science and Technology, May 2025

   Abs

   The predictive capability of a plasma discharge model depends on accurate representations of electron-impact collision cross sections, which determine the corresponding reaction rates and electron transport properties. The values of cross sections can be known only approximately either through experiments or simulations and are thus subject to uncertainties. Quantifying the uncertainties in plasma simulations allows us to assess the reliability of simulations and to provide a basis for interpreting discrepancies between simulations and experiments. For such uncertainty quantification of plasma simulations, it is essential to quantify the uncertainties of the underlying cross sections. Although much effort has been committed to calibrate the cross section values, their uncertainties are not well investigated. We characterize uncertainties in electron-argon atom collision cross sections using a Bayesian framework. Six collision processes—elastic momentum transfer, ionization, and four excitations—are characterized with semi-empirical models, which effectively capture the features important to the macroscopic properties of the plasma. A probability model for the uncertain parameters of these semi-empirical models is developed. Specifically, a Gaussian-process likelihood model is proposed to capture discrepancies among data sets, as well as the model-form inadequacies of the semiempirical models. Two other likelihood models are compared with the proposed Gaussian-process model, to illustrate the importance of the choice of the likelihood model. The cross section models are calibrated using the electron-beam experiments and ab-inito quantum simulations. The resulting calibrated uncertainties capture well the scattering among the data sets. The calibrated cross section models are further validated against swarm-parameter experiments and zero-dimensional Boltzmann equation simulations of widely used cross section datasets.

2. JCP

   A time-parallel multiple-shooting method for large-scale quantum optimal control

   N Anders Petersson , Stefanie Günther , and Seung Whan Chung

   Journal of Computational Physics, May 2025

2024

1. Nature

   Accelerating climate technologies through the science of scale-up

   Thomas Moore , Andrew A. Wong , Brian Giera , and 21 more authors

   Nature Chemical Engineering, May 2024

   Abs

   Avoiding the worst effects of climate change depends on our ability to scale and deploy technologies faster than ever before. Scale-up has largely been the domain of industrial research and development teams, but advances in modeling and experimental techniques increasingly allow early-stage researchers to contribute to the process. Here we argue that early assessments of technology market fit and how the physics governing system performance evolves with scale can de-risk technology development and accelerate deployment. We highlight tools and processes that can be used to assess both these factors at an early stage. By bringing together technical risk assessments, scaled physics modeling, data analysis and in situ experimentation within multidisciplinary teams, new technologies can be invented, developed and deployed on a shorter timetable with greater probability of success.

2. Scalable nonlinear manifold reduced order model for dynamical systems

   Ivan Zanardi , Alejandro N Diaz , Seung Whan Chung, and 2 more authors

   arXiv preprint arXiv:2412.00507, May 2024
3. Scalable physics-guided data-driven component model reduction for steady Navier-Stokes flow

   Seung Whan Chung, Youngsoo Choi , Pratanu Roy , and 6 more authors

   arXiv preprint arXiv:2410.21583, May 2024

   HTML
4. CMAME

   Divide and conquer: Learning chaotic dynamical systems with multistep penalty neural ordinary differential equations

   Dibyajyoti Chakraborty , Seung Whan Chung, Troy Arcomano , and 1 more author

   Computer Methods in Applied Mechanics and Engineering, May 2024

   Abs

   Forecasting high-dimensional dynamical systems is a fundamental challenge in various fields, such as geosciences and engineering. Neural Ordinary Differential Equations (NODEs), which combine the power of neural networks and numerical solvers, have emerged as a promising algorithm for forecasting complex nonlinear dynamical systems. However, classical techniques used for NODE training are ineffective for learning chaotic dynamical systems. In this work, we propose a novel NODE-training approach that allows for robust learning of chaotic dynamical systems. Our method addresses the challenges of non-convexity and exploding gradients associated with underlying chaotic dynamics. Training data trajectories from such systems are split into multiple, non-overlapping time windows. In addition to the deviation from the training data, the optimization loss term further penalizes the discontinuities of the predicted trajectory between the time windows. The window size is selected based on the fastest Lyapunov time scale of the system. Multi-step penalty(MP) method is first demonstrated on Lorenz equation, to illustrate how it improves the loss landscape and thereby accelerates the optimization convergence. MP method can optimize chaotic systems in a manner similar to least-squares shadowing with significantly lower computational costs. Our proposed algorithm, denoted the Multistep Penalty NODE, is applied to chaotic systems such as the Kuramoto–Sivashinsky equation, the two-dimensional Kolmogorov flow, and ERA5 reanalysis data for the atmosphere. It is observed that MP-NODE provide viable performance for such chaotic systems, not only for short-term trajectory predictions but also for invariant statistics that are hallmarks of the chaotic nature of these dynamics.

5. CMAME

   Train small, model big: Scalable physics simulators via reduced order modeling and domain decomposition

   Seung Whan Chung, Youngsoo Choi , Pratanu Roy , and 7 more authors

   Computer Methods in Applied Mechanics and Engineering, May 2024

   Abs HTML

   Numerous cutting-edge scientific technologies originate at the laboratory scale, but transitioning them to practical industry applications is a formidable challenge. Traditional pilot projects at intermediate scales are costly and time-consuming. An alternative, the pilot-scale model, relies on high-fidelity numerical simulations, but even these simulations can be computationally prohibitive at larger scales. To overcome these limitations, we propose a scalable, physics-constrained reduced order model (ROM) method. The ROM identifies critical physics modes from small-scale unit components, projecting governing equations onto these modes to create a reduced model that retains essential physics details. We also employ Discontinuous Galerkin Domain Decomposition (DG-DD) to apply ROM to unit components and interfaces, enabling the construction of large-scale global systems without data at such large scales. This method is demonstrated on the Poisson and Stokes flow equations, showing that it can solve equations about 15–40 times faster with only ∼ 1% relative error. Furthermore, ROM takes one order of magnitude less memory than the full order model, enabling larger scale predictions at a given memory limitation.

2023

1. Accelerating kinetic simulations of electrostatic plasmas with reduced-order modeling

   Ping-Hsuan Tsai , Seung Whan Chung, Debojyoti Ghosh , and 3 more authors

   arXiv preprint arXiv:2310.18493, May 2023

   HTML

2022

1. JCP

   An optimization method for chaotic turbulent flow

   Seung Whan Chung, and Jonathan B. Freund

   Journal of Computational Physics, May 2022

   Abs HTML

   Evidence indicates that quantities-of-interest in some turbulent flows can be controlled despite the overall chaotic dynamics. It is typically thought that this is via relatively deterministic, larger-scale components of the turbulence. However, finding such controls, if they exist, is challenging because chaos causes sensitivity gradients to explode and the search space to become intractably non-convex. This challenge is analyzed, and a penalty method is introduced to cope with it. In the new approach, the time domain is broken into segments approximately matching the chaos time scales, so that the solution within each segment is both physical and relatively deterministic. The initial condition of each segment is included in an adjoint-based gradient optimization, which temporarily introduces artificial ∆\mathbfq discontinuities in the overall solution. The optimization then proceeds in stages with increasing penalization of ∆\mathbfq. The method is developed and illustrated for a logistic map, the Lorenz Equation, and an advection augmented Kuramoto–Sivashinsky Equation. These examples show how the temporarily increases the search scale prior to the strong ∆\mathbfq\to0 penalization that recovers a physical solution. It is then applied to turbulent Kolmogorov flow, for which it also far outperforms a standard adjoint-based gradient search. The utility of such an optimized chaotic solution is discussed.

2020

1. JCP

   Regular sensitivity computation avoiding chaotic effects in particle-in-cell plasma methods

   Seung Whan Chung, Stephen D. Bond , Eric C. Cyr , and 1 more author

   Journal of Computational Physics, May 2020

   Abs HTML

   Particle-in-cell (PIC) simulation methods are attractive for representing species distribution functions in plasmas. However, as a model, they introduce uncertain parameters, and for quantifying their prediction uncertainty it is useful to be able to assess the sensitivity of a quantity-of-interest (QoI) to these parameters. Such sensitivity information is likewise useful for optimization. However, computing sensitivity for PIC methods is challenging due to the chaotic particle dynamics, and sensitivity techniques remain underdeveloped compared to those for Eulerian discretizations. This challenge is examined from a dual particle–continuum perspective that motivates a new sensitivity discretization. Two routes to sensitivity computation are presented and compared: a direct fully-Lagrangian particle-exact approach provides sensitivities of each particle trajectory, and a new particle-pdf discretization, which is formulated from a continuum perspective but discretized by particles to take the advantages of the same type of Lagrangian particle description leveraged by PIC methods. Since the sensitivity particles in this approach are only indirectly linked to the plasma-PIC particles, they can be positioned and weighted independently for efficiency and accuracy. The corresponding numerical algorithms are presented in mathematical detail. The advantage of the particle-pdf approach in avoiding the spurious chaotic sensitivity of the particle-exact approach is demonstrated for Debye shielding and sheath configurations. In essence, the continuum perspective makes implicit the distinctness of the particles, which circumvents the Lyapunov instability of the N-body PIC system. The cost of the particle-pdf approach is comparable to the baseline PIC simulation.