Search NASA⌕ Search

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.

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

At least 127 records · Page 7
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

Data for The Value of Reversible Carbon Storage in a Zero-Emissions World

Atmospheric carbon dioxide removal (CDR) is required to stabilize global temperature. CDR can be achieved via ecosystem-based approaches that are cost-effective but reversible (e.g., soil and forest management) or by more durable but expensive approaches (e.g., direct air capture coupled with geologic storage). Here, we examine trade-offs between these approaches, focusing on timing, climate impacts, and cost. We simulated reversible carbon accrual for a range of CDR contract structures using a general minimalist model of ecosystem carbon cycling, and parameterized it to simulate US agricultural soil management─specifically cover cropping─as a case study. We then quantified the resulting impact on atmospheric carbon and global temperature using a climate model emulator. We find that maintaining a patchwork of reversible CDR projects by replacing lapsed projects with new projects can reduce warming by 22–195 μ°C in 2100 and that the magnitude of this cooling effect depends on how effectively the patchwork is maintained. Long-term maintenance of reversible CDR projects requires institutional stability that cannot be guaranteed over multiple decades. Consequently, effective CDR ultimately requires replacing reversible projects with durable projects. To address this problem, we modeled the cost of replacing reversible agricultural soil CDR with geologic CDR. We found that using reversible CDR as a bridge to durable CDR is potentially more cost-effective as a global cooling strategy (0.20–0.81 billion USD per μ°C avoided) than perpetual maintenance of reversible CDR (0.32–1.31 billion USD per μ°C avoided) or an immediate transition to durable CDR (1.37–2.19 billion USD per μ°C avoided). However, we emphasize that institutional commitments to maintain reversible CDR projects cannot be guaranteed. Reliance on reversible CDR as a bridge to durable CDR therefore carries an unknown amount of risk and will only function if efforts to maintain reversible CDR are robust.

Carbon↗
DOE OSTI2026

Machine learning pipeline for denoising low signal-to-noise ratio and out-of-distribution transmission electron microscopy datasets

High-resolution transmission electron microscopy (HRTEM) is crucial for observing material’s structural and morphological evolution at Angstrom scales, but the electron beam can alter these processes. Devices such as CMOS-based direct-electron detectors operating in electron-counting mode can be utilized to substantially reduce the electron dosage. However, the resulting images often lead to a low signal-to-noise ratio, which requires frame integration that sacrifices temporal resolution. Several machine learning (ML) models have been recently developed to successfully denoise HRTEM images. Yet, these models are often computationally expensive, and their inference speeds on GPUs are outpaced by the imaging speed of advanced detectors, precluding in situ analysis. Furthermore, the performance of these denoising models on datasets with imaging conditions that deviate from the training datasets has not been evaluated. To mitigate these gaps, we propose a new self-supervised ML denoising pipeline specifically designed for time-series HRTEM images. This pipeline integrates a blind-spot convolution neural network with pre-processing and post-processing steps, including drift correction and low-pass filtering. Results demonstrate that our model outperforms various other ML and non-ML denoising methods in noise reduction and contrast enhancement, leading to improved visual clarity of atomic features. Additionally, the model is drastically faster than U-Net-based ML models and demonstrates excellent out-of-distribution generalization. The model’s computational inference speed is in the order of milliseconds per image, rendering it suitable for application in in-situ HRTEM experiments.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Dispersion Loss Counteracts Embedding Condensation and Improves Generalization in Small Language Models

Large language models (LLMs) achieve remarkable performance through ever-increasing parameter counts, but scaling incurs steep computational costs. To better understand LLM scaling, we study representational differences between LLMs and their smaller counterparts, with the goal of replicating the representational qualities of larger models in smaller models. We observe a geometric phenomenon which we term embedding condensation, where token embeddings collapse into a narrow cone-like subspace in some language models. Through systematic analyses across multiple Transformer families, we show that small models such as GPT2 and Qwen3-0.6B exhibit severe condensation, whereas larger models such as GPT2-x1 and Qwen3-32B are more resistant to this phenomenon. Additional observations show that embedding condensation is not reliably mitigated by knowledge distillation from larger models. To fight against it, we formulate a dispersion loss that explicitly encourages embedding dispersion during training. Experiments demonstrate that it mitigates condensation, recovers dispersion patterns seen in larger models, and yields performance gains across 10 benchmarks. We believe this work offers a principled path toward improving smaller Transformers without additional parameters.

Xiao, Xi [ORNL] (ORCID:0009000009316982)↗
DOE OSTI2025

Comparison of removal and spatial mark‐resight models for estimating wild pig density

Density estimation is critical to effectively manage invasive species and elucidate areas of highest concern. For wild pigs (Sus scrofa), the ability to estimate density is complicated because of their variable home range sizes and social structure. Common methods for estimating density (e.g., mark-recapture) may be unsuitable in management applications because additional data needs to be collected before and after management. Removal models offer a suitable alternative to estimate density changes following management and can be applied broadly across areas where management of wild pigs is ongoing. We collected wild pig removal and camera trap data from 25 private properties ranging in size from approximately 0.5 km 2 to 95 km 2 across 3 ecoregions in South Carolina, USA, from 2020–2023. We compared factors affecting consistency and precision of property-level density estimates between removal and spatial mark-resight (SMR) models. In general, excluding 1 large outlier, density estimates from removal models were between 0.60 and 15.85 wild pigs/km 2 (median = 5.34) with a median coefficient of variation (CV) of 0.76 and 95% confidence intervals for the CV between 0.70 and 0.94. Similarly, excluding 1 large outlier, density estimates from SMR were between 0.22 and 30.97 wild pigs/km 2 (median = 5.48) with a median CV of 0.39 and 95% confidence intervals for the CV between 0.38 and 1.20. We found the precision of removal models was affected primarily by the number of wild pigs dispatched in the removal period (3 months) and the ecoregion in which they were removed. None of the covariates, including the number of recaptures (a corresponding measure of sample size), influenced precision of the SMR models, although recaptures did influence the density estimates. At the individual property level, density estimates from our 2 estimators were dissimilar from each other in approximately 80% of instances, although none of the covariates we examined influenced dissimilarity. Our results provide unique insight into how sample size affects density estimates using 2 common methods and into novel SMR models that incorporate both marked and unmarked detections. In addition, the density estimates in this study can be used as a reference for wild pig densities in common land cover types throughout the southeastern United States.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2025

Stable Simulation of the Community Atmosphere Model Using Machine‐Learning Physical Parameterization Trained With Experience Replay

In recent years, machine learning (ML) models have been used to improve physical parameterizations of general circulation models (GCMs). A significant challenge of integrating ML models into GCMs is the online instability when they are coupled for long‐term simulation. We present a new strategy that demonstrates robust online stability when the physical parameterization package of an atmospheric GCM is replaced by a deep ML model. The method uses experience replay with a multistep training scheme of the ML model in which the model's own output at the previous time step is used in the training. Predicted physics tendencies in the replay buffer with the most recent errors in the training iterations are reused, making the ML model learn from its own errors. The training method reduces the gap between the offline and online environments of the ML model. The method is used to train the ML model as the physical parameterization of the Community Atmosphere Model (CAM5) with training data from the Multi‐scale Modeling Framework high resolution simulations. Three 6‐year online simulations of the CAM5 are carried out by using the ML physics package. The simulated spatial distributions of precipitation, surface temperature and zonally averaged atmospheric fields demonstrate overall better accuracy than that of the standard CAM5 and benchmark model even without the use of additional physical constraints or tuning. This work is the first to demonstrate a solution to address the online instability problem in climate modeling with ML physics by using experience replay.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Machine learning-enabled discovery of ionic liquid–solvent electrolytes exhibiting high ionic conductivity

Ionic liquids (ILs), which are a class of materials with versatile nature and growing popularity, are facing impediments toward widespread usage as electrolytes due to various factors such as low ionic conductivity, high viscosity, high market price etc. One of the ways these limitations can be addressed is by mixing ILs with a molecular solvent. In a combinatorial sense, there exists an immense number of specific IL–solvent combinations. An exhaustive experimental or even simulation-based investigation of the chemical space spanned by such combinations can be extremely time-consuming, expensive, and nearly impossible. An alternative approach is to employ machine learning-based models developed from available databases. Although there exists prior literature that integrates machine learning to investigate mixtures of specific solvents with ILs, these models lack generalization necessitating development of a large number of ML models to handle various solvents. To remedy this shortcoming, as a part of designing green electrolytes with high ionic conductivity that can have potential applications in next-generation batteries and solar cells, this work aims to develop a unified machine learning model to predict ionic conductivity of any IL–solvent mixture system. In this regard, three models, namely, Random Forest, extreme gradient boosting (XGBoost), and artificial neural network (ANN) were formulated using the NIST ILThermo database. The dataset contained 549 unique ionic liquids from 16 cation families and 81 unique solvents, representing a total of 23 712 datapoints. SHAPLEY additive explanation (SHAP) method was used to assess the impact of various features on model prediction and their significance was compared with literature to gain physical insight about the model behavior. Finally, using the developed models, approximately 2.5 million IL–solvent mixtures at five different compositions were screened at room temperature. The high-throughput screening yielded nearly 19 000 IL–solvent mixtures for which ionic conductivity was found to exceed the ionic conductivity of conventional Li-ion battery electrolyte.

25 ENERGY STORAGE↗
DOE OSTI2024

Predicting September Arctic Sea Ice: A Multimodel Seasonal Skill Comparison

This study quantifies the state of the art in the rapidly growing field of seasonal Arctic sea ice prediction. A novel multimodel dataset of retrospective seasonal predictions of September Arctic sea ice is created and analyzed, consisting of community contributions from 17 statistical models and 17 dynamical models. Prediction skill is compared over the period 2001–20 for predictions of pan-Arctic sea ice extent (SIE), regional SIE, and local sea ice concentration (SIC) initialized on 1 June, 1 July, 1 August, and 1 September. This diverse set of statistical and dynamical models can individually predict linearly detrended pan-Arctic SIE anomalies with skill, and a multimodel median prediction has correlation coefficients of 0.79, 0.86, 0.92, and 0.99 at these respective initialization times. Regional SIE predictions have similar skill to pan-Arctic predictions in the Alaskan and Siberian regions, whereas regional skill is lower in the Canadian, Atlantic, and central Arctic sectors. The skill of dynamical and statistical models is generally comparable for pan-Arctic SIE, whereas dynamical models outperform their statistical counterparts for regional and local predictions. The prediction systems are found to provide the most value added relative to basic reference forecasts in the extreme SIE years of 1996, 2007, and 2012. SIE prediction errors do not show clear trends over time, suggesting that there has been minimal change in inherent sea ice predictability over the satellite era. Overall, this study demonstrates that there are bright prospects for skillful operational predictions of September sea ice at least 3 months in advance.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2026

Modeling Multi-View Impedance-Based Cross-Geometry SOH Estimator for Li-ion Batteries

Abstract: Accurately estimating battery’s State of Health (SOH) remains challenging when models must generalize across cell designs and operating conditions. Most Electrochemical Impedance Spectroscopy (EIS)-based approaches either (i) hand-engineer a few Nyquist-plot features for shallow models—fast but does not generalize across geometries—or (ii) learn directly from Nyquist plots with deep networks, which removes manual feature extraction, yet still limited to a single plot type. As a result, cross-geometry robustness and deployability on constrained Internet of Things (IoT) devices remain open problems. We propose a compact Convolutional Neural Network (CNN) (∼ 10k parameters) that takes multi-representation EIS inputs—Nyquist (real/imaginary) and phase–magnitude (|Z|/ϕ) stacked as four channels, so the model can learn complementary degradation signatures while remaining small enough for fast inference. We build a dataset from cyclic aging of two geometries (LG INR18650MJ1 cylindrical cells and LIR2032 coin cells), acquire EIS every ten cycles from 10 kHz to 10 mHz (10 points/decade), and evaluate with leave-one-cell-out testing strategy. We further study fusion vs. single-representation inputs and assess feasibility for on-device deployment (e.g., NVIDIA Jetson device). The results show that training on multiple EIS representations improves SOH estimation accuracy and cross-geometry generalization compared to single-representation models, which uses only Nyquist or phase–magnitude plots. This design targets accurate, generalizable SOH prediction without manual feature engineering while enabling practical real-time use.

Bakr, Ahmed [The University of Alabama (UA)]↗
DOE OSTI2025

Predicting Multicomponent Aqueous Phase Adsorption Equilibria of Organic Acids Using the Generalized Brunauer–Emmett–Teller Isotherm Model

Here, to support process development of adsorptive separation of organic acids, this work presents a thermodynamic modeling methodology to predict multicomponent aqueous phase organic acid adsorption equilibria from single-component adsorption isotherms using the generalized Brunauer–Emmett–Teller isotherm model (gBET). With the organic acid fugacities rigorously accounting for the aqueous phase organic acid solution chemistry and solution nonideality, gBET precisely represents the single-component adsorption isotherms and accurately predicts the binary and ternary mixed-acid adsorption equilibria for the ranges of initial pH (∼3–7), acid concentration (100–400 mmol/L), and temperature (298.15–328.15 K) with less than 10% average absolute relative deviation. In addition, gBET with pH-independent parameters provides insights into the underlying adsorption phenomena, including the adsorbate loadings and compositions in the monolayer and subsequent layers under varying initial pH, temperature, concentration, and composition. The gBET model predictions outperform the predictions from the overloading model and the Ideal Dilute Solution Theory.

09 BIOMASS FUELS↗
DOE OSTI2025

Bootstrap current modeling in M3D-C1

Bootstrap current plays a crucial role in the equilibrium of magnetically confined plasmas, particularly in quasi-symmetric stellarators and in tokamaks, where it can represent bulk of the electric current density. Accurate modeling of this current is essential for understanding the magnetohydrodynamic (MHD) equilibrium and stability of these configurations. This study expands the modeling capabilities of M3D-C1, an extended-MHD code, by implementing self-consistent physics models for bootstrap current. It employs two analytical frameworks: a generalized Sauter model (Sauter et al. 1999 Phys. Plasmas vol. 6, no. 7, pp. 2834–2839), and a revised Sauter-like model (Redl et al. 2021 Phys. Plasmas vol. 28, no. 2, pp. 022502). The isomorphism described by Landreman et al. (2022 Phys. Rev. Lett. vol. 128, pp. 035001) is employed to apply these models to quasi-symmetric stellarators. The implementation in M3D-C1 is benchmarked against neoclassical codes, including NEO, XGCa and SFINCS, showing excellent agreement. These improvements allow M3D-C1 to self-consistently calculate the neoclassical contributions to plasma current in axisymmetric and quasi-symmetric configurations, providing a more accurate representation of the plasma behavior in these configurations. A workflow for evaluating the neoclassical transport using SFINCS with arbitrary toroidal equilibria calculated using M3D-C1 is also presented. This workflow enables a quantitative evaluation of the error in the Sauter-like model in cases that deviate from axi- or quasi-symmetry (e.g. through the development of an MHD instability).

fusion plasma↗
DOE OSTI2025

Modeling moisture outgassing from silica-filled silicones with the generalized outgassing kinetics

Even though a generalized moisture outgassing modeling has been established and published in a peer-reviewed venue for as-received silica-filled silicones, it is still not yet intuitive for use in engineering/scientific applications. A simplification of the mathematics down to a practical level and illustration of how to do it would be beneficial to the practitioners in the field. In this report, simplified and more intuitive demonstrations of how to use the generalized outgassing kinetics for the prediction of H 2 O release from as-received and previously vacuum-baked silicones will be illustrated.

36 MATERIALS SCIENCE↗
DOE OSTI2024

A flexible class of priors for orthonormal matrices with basis function-specific structure

Statistical modeling of high-dimensional matrix-valued data motivates the use of a low-rank representation that simultaneously summarizes key characteristics of the data and enables dimension reduction. Low-rank representations commonly factor the original data into the product of orthonormal basis functions and weights, where each basis function represents an independent feature of the data. However, the basis functions in these factorizations are typically computed using algorithmic methods that cannot quantify uncertainty or account for basis function correlation structure a priori. While there exist Bayesian methods that allow for a common correlation structure across basis functions, empirical examples motivate the need for basis function-specific dependence structure. We propose a prior distribution for orthonormal matrices that can explicitly model basis function-specific structure. The prior is used within a general probabilistic model for singular value decomposition to conduct posterior inference on the basis functions while accounting for measurement error and fixed effects. We discuss how the prior specification can be used for various scenarios and demonstrate favorable model properties through synthetic data examples. Finally, we apply our method to two-meter air temperature data from the Pacific Northwest, enhancing our understanding of the Earth system’s internal variability.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Effects of machine learning errors on human decision-making: manipulations of model accuracy, error types, and error importance

Abstract This study addressed the cognitive impacts of providing correct and incorrect machine learning (ML) outputs in support of an object detection task. The study consisted of five experiments that manipulated the accuracy and importance of mock ML outputs. In each of the experiments, participants were given the T and L task with T-shaped targets and L-shaped distractors. They were tasked with categorizing each image as target present or target absent. In Experiment 1, they performed this task without the aid of ML outputs. In Experiments 2–5, they were shown images with bounding boxes, representing the output of an ML model. The outputs could be correct (hits and correct rejections), or they could be erroneous (false alarms and misses). Experiment 2 manipulated the overall accuracy of these mock ML outputs. Experiment 3 manipulated the proportion of different types of errors. Experiments 4 and 5 manipulated the importance of specific types of stimuli or model errors, as well as the framing of the task in terms of human or model performance. These experiments showed that model misses were consistently harder for participants to detect than model false alarms. In general, as the model’s performance increased, human performance increased as well, but in many cases the participants were more likely to overlook model errors when the model had high accuracy overall. Warning participants to be on the lookout for specific types of model errors had very little impact on their performance. Overall, our results emphasize the importance of considering human cognition when determining what level of model performance and types of model errors are acceptable for a given task.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Dynamic Modeling of Near Isothermal Compressor for Transcritical Carbon Dioxide Cycle

Compressors are the major energy consumption components in vapor compression systems, drawing much research effort in reducing carbon emissions. The isothermal compressor integrates the compressor chamber and gas cooler to achieve near isothermal compression, reaching up to 30% energy reduction compared to the traditional isentropic compression work. This paper presents a detailed isothermal compressor model combined with a generalized liquid piston model to account for the carbon dioxide (CO2) isothermal compression process. The model is established based on MATLAB environment. The model uses the real experimental data as boundary conditions and initial settings, which also considers the CO2 solubility in liquid piston (mineral oil) for designing, optimizing and customizing the compression chambers. The validation was carried out with experimental data using a prototype with 3.5 kW capacity. The results have demonstrated the accuracy of the dynamic model (6.2% relative error for chamber pressure and 0.5 K deviation for chamber temperature), which provide a guideline for designing and customizing the isothermal compression cycle.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2025

Guiding Principles for Geochemical/Thermodynamic Model Development and Validation in Nuclear Waste Disposal: A Close Examination of Recent Thermodynamic Models for H + —Nd 3+ —NO 3 - (—Oxalate) Systems

Development of a defensible source-term model (STM), usually a thermodynamical model for radionuclide solubility calculations, is critical to a performance assessment (PA) of a geologic repository for nuclear waste disposal. Such a model is generally subjected to rigorous regulatory scrutiny. In this article, we highlight key guiding principles for STM model development and validation in nuclear waste management. We illustrate these principles by closely examining three recently developed thermodynamic models with the Pitzer formulism for aqueous H + —Nd 3+ —NO 3 - (—oxalate) systems in a reverse alphabetical order of the authors: the XW model developed by Xiong and Wang, the OWC model developed by Oakes et al., and the GLC model developed by Guignot et al., among which the XW model deals with trace activity coefficients for Nd(III), while the OWC and GLC models are for concentrated Nd(NO 3 ) 3 electrolyte solutions. The principles highlighted include the following: (1) Principle 1. Validation against independent experimental data: A model should be validated against experimental data or field observations that have not been used in the original model parameterization. We tested the XW model against multiple independent experimental data sets including electromotive force (EMF), solubility, water vapor, and water activity measurements. The results show that the XW model is accurate and valid for its intended use for predicting trace activity coefficients and therefore Nd solubility in repository environments. (2) Principle 2. Testing for relevant and sensitive variables: Solution pH is such a variable for an STM and easily acquirable. All three models are checked for their ability to predict pH conditions in Nd(NO 3 ) 3 electrolyte solutions. The OWC model fails to provide a reasonable estimate for solution pH conditions, thus casting serious doubt on its validity for a source-term calculation. In contrast, both the XW and GLC models predict close-to-neutral pH values, in agreement with experimental measurements. (3) Principle 3. Honoring physical constraints: Upon close examination, it is found that the Nd(III)-NO 3 association schema in the OWC model suffers from two shortcomings. Firstly, its second stepwise stability constant for Nd(NO 3 ) 2+ (log K 2 ) is much higher than the first stepwise stability constant for NdNO 3 2+ (log K 1 ), thus violating the general rule of (log K 2 –log K 1 ) < 0, or $\frac{K1}{K2}$>1. Secondly, the OWC model predicts abnormally high activity coefficients for Nd(NO 3 ) 2 + (up to ~900) as the concentration increases. (4) Principle 4. Minimizing degrees of freedom for model fitting: The OWC model with nine fitted parameters is compared with the GLC model with five fitted parameters, as both models apply to the concentrated region for Nd(NO 3 ) 3 electrolyte solutions. The latter appears superior to the former because the latter can fit osmotic coefficient data equally well with fewer model parameters. The work presented here thus illustrates the salient points of geochemical model development, selection, and validation in nuclear waste management.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2025

Continued Testing and Modeling of the AirSquared All-Metal Tritium Scroll Pump

The implementation of fusion energy requires the processing of the deuterium-tritium mixture used to fuel the reaction. Many gas processing operations require either a gas-motive force or a pressure gradient across a system; both are achieved using vacuum pumps. However, because of the material challenges of working with tritium, specialized pumps are required that use only metal on wetted surfaces. This prevents tritium retention and degradation of materials such as oils and polymers and reduces the generation of tritiated waste. Typical gas handling systems for tritium operations involve metal bellows pumps, all-metal scroll pumps, and metal turbomolecular pumps, depending on the pressure requirements. With the industry standard Normetex Type 15 m 3 /h (typically referred to as the Normetex 15) pump no longer available, the fusion market requires viable alternatives. In this study, we test an American-made pump, the AirSquared V16H034A-C01, and compare its performance to the Normetex. A performance model of the AirSquared pump is also produced, which can be applied to general process models to generate an accurate description of pressures and flows through various unit operations.

08 HYDROGEN↗