SEARCH · Search NASA

Results for “Model generalization”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 361 records · Page 20

All sources NASA NTRS DOE OSTI

DOE OSTI2025

Depinning, melting, and sliding of generalized Wigner crystals in Moiré systems

We numerically examine the depinning, sliding, and melting of commensurate and incommensurate Wigner crystals on two-dimensional hexagonal periodic substrates near fillings of 1 / 3 , 1 / 2 , and 2 / 3 to model the dynamics of generalized Wigner crystals in moiré heterostructures. At low temperatures where thermal fluctuations are irrelevant, for commensurate fillings we find a strongly pinned state that depins into a sliding crystal, while at incommensurate fillings the depinning threshold is strongly reduced. For fillings below 1 / 2 , the depinning occurs in a two-step process. Above the first depinning threshold, there is an extended range of drives where the conduction occurs via the sliding of antikinks along the charge stripes, followed by a second threshold where all the charges begin to slide. At finite temperatures, the external driving reduces the effective melting temperature at commensurate fillings and enhances creep at incommensurate fillings. We show that depinning into different sliding states, such as moving fluids or moving crystals, produces nonlinear features in the transport curves. We also show that transport is asymmetric on either side of a commensurate filling due to the different dynamics for interstitials versus holes in the commensurate structure. A variety of sliding states should be accessible for generalized Wigner crystals at finite temperatures even for low drives. If experiments can realize sliding dynamics in moiré systems, it would open another class of commensurate and incommensurate phases for study. Additionally, the sliding dynamics and antikink flow could provide unique functionalities for charge transport in moiré heterostructures. Published by the American Physical Society 2025

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

DOE OSTI2024

Chiral-odd GPDs in the bag model

A study of chiral-odd generalized parton distributions (GPDs) of the nucleon is presented in the bag model demonstrating that in this model all four chiral-odd GPDs are nonzero contrary to other claims in literature. The bag model results for the GPDs H T q ( x , ξ , t ) , E T q ( x , ξ , t ) , H ˜ T q ( x , ξ , t ) agree with other models within a typical quark model accuracy. We present one of the few quark model calculations where polynomiality is satisfied and the sum rule ∫ d x E ˜ T q ( x , ξ , t ) = 0 holds. We confront our results with predictions from the large- N c limit, and with lattice QCD calculations. We conclude that the bag model successfully catches the main features of chiral-odd GPDs. Published by the American Physical Society 2024

Tezgin, Kemal (ORCID:0000000194929512)↗

DOE OSTI2026

ssys

`ssys` is a Python toolkit for exact algebraic transformation of ordinary differential equation (ODE) models into canonical S-system or Generalized Mass Action (GMA) form. Given a model in Antimony or SBML format, `ssys` produces a mathematically equivalent representation. The transformation introduces auxiliary variables as needed to decompose a broad class of nonlinearities into products of power-law terms. The recast is exact: the original and transformed systems have identical dynamics on the invariant constraint manifold defined by auxiliary variable definitions, given consistent initial conditions.

Hlavacek, William [Los Alamos National Laboratory]↗

DOE OSTI2024

A modular and extensible CHARMM-compatible model for all-atom simulation of polypeptoids

Peptoids (N-substituted glycines) are a class of sequence-defined synthetic peptidomimetic polymers with applications including drug delivery, catalysis, and biomimicry. Classical molecular simulations have been used to predict and understand the conformational dynamics of single chains and their self-assembly into morphologies including sheets, tubes, spheres, and fibrils. The CGenFF-NTOID model based on the CHARMM General Force Field has demonstrated success in accurate all-atom molecular modeling of peptoid structure and thermodynamics. Extension of this force field to new peptoid side chains has historically required reparameterization of side chain bonded interactions against ab initio data. This fitting protocol improves the accuracy of the force field but is also burdensome and precludes modular extensibility of the model to arbitrary peptoid sequences. In this work, we develop and demonstrate a Modular Side Chain CGenFF-NTOID (MoSiC-CGenFF-NTOID) as an extension of CGenFF-NTOID employing a modular decomposition of the peptoid backbone and side chain parameterizations, wherein arbitrary side chains within the large family of substituted methyl groups (i.e., –CH 3 , –CH 2 R, –CHRR', and –CRR'R") are directly ported from CGenFF. We validate this approach against ab initio calculations and experimental data to develop a MoSiC-CGenFF-NTOID model for all 20 natural amino acid side chains along with 13 commonly used synthetic side chains and present an extensible paradigm to efficiently determine whether a novel side chain can be directly incorporated into the model or whether refitting of the CGenFF parameters is warranted. We make the model freely available to the community along with a tool to perform automated initial structure generation.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2025

Many-body expansion based machine learning models for octahedral transition metal complexes

Abstract Graph-based machine learning (ML) models for material properties show great potential to accelerate virtual high-throughput screening of large chemical spaces. However, in their simplest forms, graph-based models do not include any 3D information and are unable to distinguish stereoisomers such as those arising from different orderings of ligands around a metal center in coordination complexes. In this work we present a modification to revised autocorrelation descriptors, a molecular graph featurization method, for predicting spin state dependent properties of octahedral transition metal complexes (TMCs). Inspired by analytical semi-empirical models for TMCs, the new modeling strategy is based on the many-body expansion (MBE) and allows one to tune the captured stereoisomer information by changing the truncation order of the MBE. We present the necessary modifications to include this approach in two commonly used ML methods, kernel ridge regression and feed-forward neural networks. On a test set composed of all possible isomers of binary TMCs, the best MBE models achieve mean absolute errors (MAEs) of 2.75 kcal mol −1 on spin-splitting energies and 0.26 eV on frontier orbital energy gaps, a 30%–40% reduction in error compared to models based on our previous approach. We also observe improved generalization to previously unseen ligands where the best-performing models exhibit MAEs of 4.00 kcal mol −1 (i.e. a 0.73 kcal mol −1 reduction) on the spin-splitting energies and 0.53 eV (i.e. a 0.10 eV reduction) on the frontier orbital energy gaps. Because the new approach incorporates insights from electronic structure theory, such as ligand additivity relationships, these models exhibit systematic generalization from homoleptic to heteroleptic complexes, allowing for efficient screening of TMC search spaces.

Meyer, Ralf (ORCID:0000000322360261)↗

DOE OSTI2026

Site-decorated model for unconventional frustrated magnets: Ultranarrow phase crossover and two-dimensional spin reversal transition

Here, the site-decorated Ising model is introduced to advance the understanding and experimental realization of the recently discovered one-dimensional (1D) finite-temperature ultranarrow phase crossover in an external magnetic field, while mitigating the geometric complexities of traditional bond-decorated models. The unconventional frustration and physics are clarified by exactly mapping the 1D site-decorated Ising model in a magnetic field onto a zero-field bond-decorated 𝐽 1 −𝐽 2 Ising model with conventional geometrical frustration. Furthermore, although higher-dimensional Ising models in an external field remain unsolved exactly, an exact solution for a spin-reversal transition—driven by an exotic, hidden half-ice, half-fire state induced by site decoration—is derived. This transition, triggered by a slight variation in temperature or magnetic field—without changing its direction—even in the weak-field limit, offers a promising route toward energy-efficient applications such as data storage and processing. The results suggest that site decoration offers an avenue for materials and device design, particularly in systems such as mixed 𝑑−𝑓 compounds, optical lattices, and neural networks, calling for further studies with site-decorated Heisenberg models. In addition, the site-decorated model offers a rigorous test ground for artificial intelligence (AI) in science, as the analytic derivation of the present results was not only validated but also improved by a general-purpose large language model, inspiring the use of AI as scientific discoverer.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

DOE OSTI2024

Model-independent description of B → D π ℓ ν decays

We introduce a new parametrization of B → D π ℓ ν form factors using a partial-wave expansion and derive bounds on the series coefficients using analyticity and unitarity. This is the first generalization of the model-independent formalism developed by Boyd, Grinstein, and Lebed for B → D ℓ ν to semileptonic decays with multihadron final states, and enables data-driven form-factor determinations with robust, systematically improvable uncertainties. Using this formalism, we extract the form-factor parameters for B → D 2 * ( → D π ) ℓ ν decays in a model-independent way from fits of data from the Belle Experiment. We find that the semileptonic data are compatible with the presence of two poles in the D π S-wave channel, which is the scenario preferred by nonleptonic decays and unitarized chiral perturbation theory. Published by the American Physical Society 2024

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

DOE OSTI2025

Physics-informed machine learning for building performance simulation-A review of a nascent field

Building performance simulation (BPS) is critical for understanding building dynamics and behavior, analyzing the performance of the built environment, optimizing energy efficiency, improving demand flexibility, and enhancing building resilience. However, conducting BPS is not trivial. Traditional BPS relies on accurate building energy models, which are primarily physics-based and heavily dependent on detailed building information, expert knowledge, and case-by-case model calibrations, significantly limiting their scalability. With the development of sensing technology and the increased availability of data, there is growing attention and interest in data-driven BPS. However, purely data-driven models often suffer from limited generalization ability and a lack of physical consistency, resulting in poor performance in real-world applications. To address these limitations, recent studies have begun integrating physics priors into data-driven models, a methodology known as physics-informed machine learning (PIML). PIML is an emerging field where its definitions, methodologies, evaluation criteria, application scenarios, and future directions remain open. To bridge those gaps, this study systematically reviews the state-of-the-art PIML for BPS, offering a comprehensive definition of PIML and comparing it to traditional BPS approaches regarding data requirements, modeling effort, performance, and computational cost. We also summarize the commonly used methodologies, validation approaches, application domains, available data sources, open-source packages, and testbeds. In addition, this study provides a general guideline for selecting appropriate PIML models based on BPS applications. Finally, this study identifies key challenges and outlines future research directions, providing a solid foundation and valuable insights to advance R&D of PIML in BPS.

Jiang, Zixin↗

DOE OSTI2026

Dataset for Blueprinting Electrified Transit System Implementation

This dataset contains the figures and tabulated results generated from a system-level optimization study of transit fleet electrification planning. The dataset does not include executable modeling code required to reproduce the optimization. The dataset includes results for optimized charging infrastructure deployment by location and power level and service block assignments by fuel type, battery capacity selections, and distributed energy resource sizing. It also contains aggregated financial results, capital expenditures, operating cost summaries, net present cost comparisons across scenarios, and quantified air quality impacts. Results are structured to reflect multiple planning scenarios, including heuristic electrification plans, system-optimized configurations, and sensitivity cases with alternative objective weightings. The modeling was developed using publicly available General Transit Feed Specification data from Omnitrans and standardized modeling assumptions.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

DOE OSTI2026

Dataset for Blueprinting Electrified Transit System Implementation

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

DOE OSTI2025

Hyper Spectral Anomaly Detection

Anomaly detection is a common machine learning (ML) task with growing importance in the fields of imaging, quality assurance, and multiple security related disciplines. Anomaly detection is more difficult than traditional machine learning methods due to the inherent unlabeled nature of the datasets. Existing anomaly detection architectures commonly face challenges with explainability, retaining information related to the relational structure of the data, and false positive rates. Hyperspectral Imaging Anomaly Detection (HSI) is a statistical model that employs vertex and edge weighted graphs to preserve the data’s relationships on different topographical scales. The model is able to generalize from anomaly detection in 2D images to novel datasets related to cyber-security. Furthermore, the use of multi-spectral and other filtering methods results in fewer false positives and increases the explainability of model predictions. When applying HSI to cyber-security datasets, we are able to successfully detect malicious activity with a relatively high degree of accuracy.

97 - MATHEMATICS AND COMPUTING↗

DOE OSTI2025

Comparing computational times for simulations when using PBPK model template and stand-alone implementations of PBPK models

Introduction We previously developed a PBPK model template that consists of a single model “superstructure” with equations and logic found in many physiologically based pharmacokinetic (PBPK) models. Using the template, one can implement PBPK models with different combinations of structures and features. Methods To identify factors that influence computational time required for PBPK model simulations, we conducted timing experiments using various implementations of PBPK models for dichloromethane and chloroform, including template and stand-alone implementations, and simulating four different exposure scenarios. For each experiment, we measured the required computational time and evaluated the impacts of including various model features (e.g., number of output variables calculated) and incorporating various design choices (e.g., different methods for estimating blood concentrations). Results We observed that model implementations that treat body weight and dependent quantities as constant (fixed) parameters can result in a 30% time savings compared with options that treat body weight and dependent quantities as time-varying. We also observed that decreasing the number of state variables by 36% in our PBPK model template led to a decrease of 20–35% in computational time. Other factors, such as the number of output variables, the method for implementing conditional statements, and the method for estimating blood concentrations, did not have large impacts on simulation time. In general, simulations with PBPK model template implementations of models required more time than simulations with stand-alone implementations, but the flexibility and (human) time savings in preparing and reviewing a model implemented using the PBPK model template may justify the increases in computational time requirements. Conclusion Our findings concerning how PBPK model design and implementation decisions impact computational speed can benefit anyone seeking to develop, improve, or apply a PBPK model, with or without the PBPK model template.

Bernstein, Amanda S.↗

DOE OSTI2025

Realizability-preserving discontinuous Galerkin method for spectral two-moment radiation transport in special relativity

Here we present a realizability-preserving numerical method for solving a spectral two-moment model to simulate the transport of massless, neutral particles interacting with a steady background material moving with relativistic velocities. The model is obtained as the special relativistic limit of a four-momentum-conservative general relativistic two-moment model. Using a maximum-entropy closure, we solve for the Eulerian-frame energy and momentum. The proposed numerical method is designed to preserve moment realizability, which corresponds to moments defined by a nonnegative phase-space density. The realizability-preserving method is achieved with the following key components: (i) a discontinuous Galerkin phase-space discretization with specially constructed numerical fluxes in the spatial and energy dimensions; (ii) a strong stability-preserving implicit-explicit time-integration method; (iii) a realizability-preserving conserved to primitive moment solver; (iv) a realizability-preserving implicit collision solver; and (v) a realizability-enforcing limiter. Component (iii) is necessitated by the closure procedure, which closes higher order moments nonlinearly in terms of primitive moments. The nonlinear conserved to primitive and the implicit collision solves are formulated as fixed-point problems, which are solved with custom iterative solvers designed to preserve the realizability of each iterate. With a series of numerical tests, we demonstrate the accuracy and robustness of this discontinuous-Galerkin-implicit-explicit method.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2025

Learning broken symmetries with approximate invariance

Recognizing symmetries in data allows for significant boosts in neural network training, which is especially important where training data are limited. In many cases, however, the exact underlying symmetry is present only in an idealized dataset, and is broken in actual data, due to asymmetries in the detector, or varying response resolution as a function of particle momentum. Standard approaches, such as data augmentation or equivariant networks fail to represent the nature of the full, broken symmetry, effectively overconstraining the response of the neural network. We propose a learning model which balances the generality and asymptotic performance of unconstrained networks with the rapid learning of constrained networks. This is achieved through a dual-subnet structure, where one network is constrained by the symmetry and the other is not, along with a learned symmetry factor. In a simplified toy example that demonstrates violation of Lorentz invariance, our model learns as rapidly as symmetry constrained networks but escapes its performance limitations.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DOE OSTI2025

Efficient Training of Deep Neural Operator Networks via Randomized Sampling

Neural operators (NOs) employ deep neural networks to learn the mappings between infinitedimensional function spaces. Deep operator network (DeepONet), a popular NO architecture, has demonstrated success in the real-time prediction of complex dynamics across various scientific and engineering applications. In this work, we introduce a random sampling technique to be adopted during the training of DeepONet, aimed at improving the generalization ability of the model, while significantly reducing the computational time. The proposed approach targets the trunk network of the DeepONet model that outputs the basis functions corresponding to the spatiotemporal locations of the bounded domain on which the physical system is defined. While constructing the loss function, DeepONet training traditionally considers a uniform grid of spatiotemporal points at which all the output functions are evaluated for each iteration. This approach leads to a larger batch size, resulting in poor generalization and increased memory demands, due to the limitations of the stochastic gradient descent (SGD) optimizer. The proposed random sampling over the inputs of the trunk net mitigates these challenges, improving generalization and reducing the memory requirements during training, resulting in significant computational gains. We validate our hypothesis through three benchmark examples, demonstrating substantial reductions in training time while achieving comparable or lower overall test errors relative to the traditional training approach. Our results indicate that incorporating randomization in the trunk network inputs during training enhances the efficiency and robustness of DeepONet, offering a promising avenue for improving the framework’s performance in modeling complex physical systems.

Karumuri, Sharmila [Department of Civil & Systems ↗

DOE OSTI2025

A foundation model for atomistic materials chemistry

Atomistic simulations of matter, especially those that leverage first-principles (ab initio) electronic structure theory, provide a microscopic view of the world, underpinning much of our understanding of chemistry and materials science. Over the last decade or so, machine-learned force fields have transformed atomistic modeling by enabling simulations of ab initio quality over unprecedented time and length scales. However, early machine-learning (ML) force fields have largely been limited by (i) the substantial computational and human effort required to develop and validate potentials for each particular system of interest and (ii) a general lack of transferability from one chemical system to the next. Here, we show that it is possible to create a general-purpose atomistic ML model, trained on a public dataset of moderate size, that is capable of running stable molecular dynamics for a wide range of molecules and materials. We demonstrate the power of the MACE-MP-0 model-and its qualitative and at times quantitative accuracy-on a diverse set of problems in the physical sciences, including properties of solids, liquids, gases, chemical reactions, interfaces, and even the dynamics of a small protein. The model can be applied out of the box as a starting or "foundation" model for any atomistic system of interest and, when desired, can be fine-tuned on just a handful of application-specific data points to reach ab initio accuracy. Establishing that a stable force-field model can cover almost all materials changes atomistic modeling in a fundamental way: experienced users obtain reliable results much faster, and beginners face a lower barrier to entry. Foundation models thus represent a step toward democratizing the revolution in atomic-scale modeling that has been brought about by ML force fields.

Batatia, Ilyes↗

DOE OSTI2024

Pretraining Billion-Scale Geospatial Foundational Models on Frontier

As AI workloads increase in scope, generalization capability becomes challenging for small task-specific models and their demand for large amounts of labeled training samples increases. On the contrary, Foundation Models (FMs) are trained with internet-scale unlabeled data via self-supervised learning and have been shown to adapt to various tasks with minimal fine-tuning. Although large FMs have demonstrated significant impact in natural language processing and computer vision, efforts toward FMs for geospatial applications have been restricted to smaller size models, as pretraining larger models requires very large computing resources equipped with state-of-the-art hardware accelerators. Current satellite constellations collect 100+TBs of data a day, resulting in images that are billions of pixels and multimodal in nature. Such geospatial data poses unique challenges opening up new opportunities to develop FMs. We investigate billion scale FMs and HPC training profiles for geospatial applications by pretraining on publicly available data. We studied from end-to-end the performance and impact in the solution by scaling the model size. Our larger 3B parameter size model achieves up to 30% improvement in top1 scene classification accuracy when comparing a 100M parameter model. Moreover, we detail performance experiments on the Frontier supercomputer, America's first exascale system, where we study different model and data parallel approaches using PyTorch's Fully Sharded Data Parallel library. Specifically, we study variants of the Vision Transformer architecture (ViT), conducting performance analysis for ViT models with size up to 15B parameters. By discussing throughput and performance bottlenecks under different parallelism configurations, we offer insights on how to leverage such leadership-class HPC resources when developing large models for geospatial imagery applications.

Tsaris, Aristeidis (aris)↗

DOE OSTI2024

Force Field X: A computational microscope to study genetic variation and organic crystals using theory and experiment

Force Field X (FFX) is an open-source software package for atomic resolution modeling of genetic variants and organic crystals that leverages advanced potential energy functions and experimental data. FFX currently consists of nine modular packages with novel algorithms that include global optimization via a many-body expansion, acid–base chemistry using polarizable constant-pH molecular dynamics, estimation of free energy differences, generalized Kirkwood implicit solvent models, and many more. Applications of FFX focus on the use and development of a crystal structure prediction pipeline, biomolecular structure refinement against experimental datasets, and estimation of the thermodynamic effects of genetic variants on both proteins and nucleic acids. The use of Parallel Java and OpenMM combines to offer shared memory, message passing, and graphics processing unit parallelization for high performance simulations. Overall, the FFX platform serves as a computational microscope to study systems ranging from organic crystals to solvated biomolecular systems.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗