Search NASA⌕ Search

SEARCH · Search NASA

Results for “Pre-trained models”

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 37 records · Page 2
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Autonomous phase mapping of gold nanoparticles synthesis with differentiable models of spectral shape

Autonomous experimentation–or self-driving labs–offers a systematic approach to accelerate materials discovery by integrating automated synthesis, characterization, and data-driven decision-making. We present a closed-loop workflow for the on-demand synthesis and structural characterization of colloidal gold nanoparticles, enabling direct mapping from composition to nanoscale structure. Our framework leverages differentiable models of spectral shape to address two central tasks in self-driving labs: (a) phase mapping, or identifying compositional regions with distinct structural behavior; and (b) material retrosynthesis, or optimizing compositions for target structure. Using functional data analysis, we develop a data-driven model with generative pre-training, active learning, and high-throughput experiments to predict spectral responses across composition space. We demonstrate the approach on seed-mediated growth of gold nanoparticles, showcasing its ability to extract design rules, reveal secondary interactions, and efficiently navigate morphology space. Gradient-based optimization of the models enables inverse design, making this a unified platform.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Leveraging large language models to address data scarcity in machine learning for graphene synthesis

Machine learning in experimental materials science faces significant challenges due to the scarcity of data, which are costly and time-consuming to generate, particularly when relying on in-house experiments. Literature data mining offers a potential solution but introduces issues like mixed data quality, inconsistent formats, and non-uniform reporting of synthesis parameters, resulting in partially missing and heterogeneous features across the dataset. Here, we propose data imputation and feature engineering methods that employ pre-trained large language models (LLMs) to enhance machine learning performance on scarce, heterogeneous datasets, demonstrated on graphene CVD synthesis data and the ML-HydPARK hydrogen storage dataset. GPT models perform data imputation via tailored prompting and semantic normalization of inconsistently reported features through embeddings, for example, to harmonize the complex nomenclature of CVD substrates. Beyond yielding more diverse and richer feature representations than traditional methods such as K-nearest neighbors (KNN) and Multivariate Imputation by Chained Equations (MICE), LLM-based data imputation is evaluated against dataset characteristics and prompting strategies. We vary the level of autonomy granted to the LLM, from generic prompting that leverages pre-trained knowledge for autonomous data generation to data-informed prompting that constrains outputs using target-specific information, and demonstrate which level of autonomy yields superior imputation performance across datasets and feature types. The proposed data engineering methods markedly improve downstream performance; for example, in graphene layer number classification using a support vector machine (SVM), binary accuracy increases from 39% to 65% and ternary accuracy from 52% to 72%. Fine-tuning experiments on both datasets show that combining our proposed LLM-based data imputation and feature encoding methods with numerical machine learning predictors outperforms standalone fine-tuned LLM predictors in data-scarce settings. The proposed strategies emphasize data enhancement techniques rather than refining learning architectures or regularizing loss functions, offering a broadly applicable framework for improving machine learning performance on scarce, inhomogeneous datasets.

Chemical vapor deposition↗
DOE OSTI2025

A Representation Fusion Framework for Decoupling Diagnostic Information in Multimodal Learning

Modern medicine increasingly relies on multimodal data, ranging from clinical notes to imaging and genomics, to guide diagnosis and treatment. However, integrating these heterogeneous data sources in a principled and interpretable manner remains a major challenge. We present MODES (Multi-mOdal Disentangled Embedding Space), a representation fusion framework that explicitly separates shared and modality-specific factors of variation, offering a structured latent space for multimodal information that improves both prediction and interpretability. By leveraging pre-trained unimodal foundation models, MODES mitigates the dependency on extensive paired datasets, crucial in data-scarce clinical settings. We introduce a masking strategy that optimizes representation dimensionality by eliminating low-information dimensions, to achieve compact, information-rich representations. Our framework demonstrates superior performance in predicting diagnoses and phenotypes compared to unimodal and conventional fusion models. MODES also enables robust diagnostic inference in missing data scenarios, offering an opportunity toward interpretable and efficient multimodal diagnostics in personalized healthcare.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2025

RingX: Scalable Parallel Attention for Long-Context Learning on HPC

The attention mechanism has become foundational for remarkable AI breakthroughs since the introduction of the Transformer, driving the demand for increasingly longer context to power frontier models such as large-scale reasoning language models and high-resolution image/video generators. However, its quadratic computational and memory complexities present substantial challenges. Current state-of-the-art parallel attention methods, such as ring attention, are widely adopted for long-context training but utilize a point-to-point communication strategy that fails to fully exploit the capabilities of modern HPC network architectures. In this work, we propose ringX, a scalable family of parallel attention methods optimized explicitly for HPC systems. By enhancing workload partitioning, refining communication patterns, and improving load balancing, ringX achieves up to 3.4 × speedup compared to conventional ring attention on the Frontier supercomputer. Optimized for both bi-directional and causal attention mechanisms, ringX demonstrates its effectiveness through training benchmarks of a Vision Transformer (ViT) on a climate dataset and a Generative Pre-Trained Transformer (GPT) model, Llama3 8B. Our method attains an end-to-end training speedup of approximately 1.5 × in both scenarios. To our knowledge, the achieved 38% model FLOPs utilization (MFU) for training Llama3 8B with a 1M-token sequence length on 4,096 GPUs represents one of the highest training efficiencies reported for long-context learning on HPC systems. Our code implementation is available at https://github.com/jqyin/ringX-attention.

Yin, Junqi [ORNL] (ORCID:0000000338435520)↗
DOE OSTI2025

Activation Domain Hunter (ADhunter) v2.0

ADhunter is a software program that enables accurate identification and quantification of transcriptional activation domains. Unlike previous software, ADhunter uses protein representations from a pre-trained protein language model, model ensembling, and a training dataset from a diverse sampling of protein sequence space for state-of-the-art performance. These advantages enable improved perception of transcriptional activation domains across sequence space that can be used for mapping natural genetic circuits and engineering synthetic genetic circuits. In particular, ADhunter enables fine-tuned control of gene expression through synthetic transcription factors that can be used for complex control of cellular programs.

Waldburger, Lucas [Lawrence Berkeley National Labo↗
DOE OSTI2026

Neural Scaling Laws for Jet Generation

Recently observed empirical scaling laws describe the performance of foundation-type models as three independent key quantities -- dataset size, compute, and model parameters -- are modified. Extracting these scaling laws informs the training of large complex models for which the tuning of hyperparameters in traditional ways is not feasible. This work for the first time explores if scaling laws can also be observed for the task of particle jet generation -- both relevant as a pre-training objective for foundation models and as in-situ simulation by itself. We indeed replicate the key logarithmic scaling law behavior for model-size scaling. Beyond studying the next token prediction validation loss of the generative model, we also study the sliced Wasserstein distance of five physical quantities that are not immediately available to the model during training. Our study shows that this quantity is monotonically related to the next token prediction validation loss, meaning that this loss is indeed a good proxy for the physics performance. For the scaling with dataset size and compute, we observe substantially weaker scaling behavior of both the loss and the sliced Wasserstein distance. We analyze this behavior by introducing the concept of a learnable window, and argue that autoregressive next token prediction on jet constituents exhibits comparatively rapid saturation relative to language-model studies. We discuss possible origins of this behavior, including the stochastic nature of QCD radiation and differences between generative and supervised learning tasks in collider physics.

Amram, Oz [Fermilab]↗
DOE OSTI2026

Enhancing transfer learning in angle-resolved photoemission spectroscopy (ARPES) with spatially-aware representations via graph convolution

A recent application of machine learning has been to spatially-resolved angle-resolved photoemission spectroscopy (ARPES). Here we advance the state-of-the-art by applying representational learning to transform ARPES data into an embedding space of a pre-trained self-supervised learning model, thus enhancing the pipeline that improves the bandstructure classification and domain assignment/segmentation performance compared to a k-means clustering method. In the current iteration, the real-space information is entered into the domain assignment through the graph convolution method, which improves the transfer learning performance of the original self-supervised model. Lastly, an unsupervised automated tool is developed that incorporates these techniques to enable automatic domain assignment.

ARPES↗
DOE OSTI2025

Transfer learning-based soybean LAI estimations by integrating PROSAIL, UAV, and PlanetScope imagery

Accurate Leaf Area Index (LAI) estimations at the soybean plot scale is achievable using high-resolution Unmanned Aerial Vehicle (UAV) imagery and field measurement samples. However, the limited coverage of UAV flights restricts large-scale remote sensing monitoring in expansive soybean fields. This study leverages the broad coverage and 3-m resolution of PlanetScope satellite imagery to extend LAI prediction from UAV to satellite scales through transfer learning, using UAV-scale LAI estimates as a benchmark to validate cross-scale consistency. To address this challenge, this study proposed the LAI-TransNet, a two-stage transfer learning framework designed for precise and scalable soybean LAI prediction across large areas, demonstrating its effectiveness in cross-scale monitoring. In Stage 1, a UAV-scale benchmark is established using PROSAIL-simulated UAV reflectance data (UAV-Sim) and field-measured soybean LAI. Traditional machine learning, deep learning, and transfer learning models are trained on a hybrid UAV-Sim and field-measured dataset (UAV-Sim_Measured), with the transfer learning model CNN-TL, fine-tuned using pre-trained weights derived from UAV-Sim, achieving the highest accuracy (R 2 = 0.81, RMSE = 0.64 m 2 /m 2 , rRMSE = 11.5 %). In Stage 2, LAI-TransNet is developed by fine-tuning the CNN-TL model on PlanetScope simulated data (PS-Sim), preprocessed via cross-domain mapping to align UAV and satellite spectral features. Real PlanetScope imagery is corrected for reflectance consistency with reference to UAV imagery spectral profiles. LAI-TransNet outperforms other deep learning models trained directly on PS-Sim (R 2 = 0.69 vs. 0.60–0.63), ensuring robust cross-scale consistency. In conclusion, by bridging UAV and satellite scales, LAI-TransNet enables large-scale soybean LAI monitoring, enhancing precision agriculture management through improved monitoring with the PlanetScope imagery.

Leaf area index (LAI)↗
DOE OSTI2025

Derivation of physical equations for high-speed laser welding using large language models

It is challenging to formulate complex physical phenomena that occur in a manufacturing process, particularly when the available data are limited, rendering conventional data-driven approaches ineffective. This study aims to predict humping onset in high-speed laser welding by introducing a novel framework, namely text-to-equations generative pre-trained transformer (T2EGPT). This method leverages the capabilities of large language models (LLMs), in combination with sparse experimental data and enriched literature data, to derive an interpretable and generalizable equation for predicting humping initiation. By capturing key correlations among physical parameters, T2EGPT generates a compact and dimensionless expression that accurately predicts hump formation. The equation reveals that humping arises from the interplay between inertia-driven backward melt flow and capillary-driven surface stabilization, where inertial forces drive molten metal backward and capillary forces resist surface deformation. Furthermore, compared to traditional data-driven models, T2EGPT demonstrates enhanced predictive accuracy and cross-material transferability. More broadly, this study highlights the potential of LLMs to integrate textual information with data-driven discovery, enabling the extraction of physical laws in data-scarce scientific domains.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Structure-Aware Unsupervised, Transformational Machine Learning for Drug Discovery (DTRA Basic Research Final Report)

The major goal of this project is to develop machine learning (ML) methods to enable improved predictive power on real drug discovery for novel targets. More specifically, we planned to demonstrate the capability and effectiveness of ML tools utilizing unlabeled large-volume protein-ligand datasets. We investigated multiple pre-training approaches for 3D protein-ligand structure-based foundation models, without relying on experimental binding data. We also addressed scenarios in which crystal structures are unavailable or binding data are limited. We also planned to develop a complete pipeline to screen novel compounds as well as to demonstrate the capability and effectiveness of the developed methods by testing on a realistic drug discovery task such as SARS-CoV-2. While the major goals and milestones remain consistent with the original proposal, certain technical details have been adjusted, based on the experimental results and related outcomes.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

ChatPORT: Fine-Tuned LLM for Easy Code {PORT}ing

Fine-tuning existing LLMs for specialized tasks has become a very attractive alternative due to its low cost and quick development cycle. With many pre-trained LLMs available, it is an increasingly complex task to choose the correct model as the starting point or base model. In this work we discuss ChatPORT - a specialized fine-tuned LLM geared towards providing correctly translated codes from one programming model to another. We evaluate a number of base models and compare and contrast their features and characteristics that make them a viable starting point. In this paper, we focus on the OpenMP offload porting capabilities of ChatPORT. We build our training data using kernels from the Heterogeneous Computing Benchmarks (HeCBench) [12] and the OpenMP Validation and Verification suite [5] to fine-tune the base models. We then test the model using unseen kernels extracted from the HeCBench benchmark suite. Our results show that: (1) not all open LLMs geared towards HPC are aware of programming models like OpenMP, (2) although all base models benefit from fine-tuning they learn differently and produce different correctness rates, (3) depending on the memory size and compute resource available, different base models can be used for fine-tuning without significantly affecting the quality of transpiled code they generate, (4) fine-tuning improved the correctness rate of the LLM by an average of 43.2%, and (5) feedback-based training data further increased the correctness rate by an average of 6% over the LLMs tested.

Pophale, Swaroop [ORNL] (ORCID:0000000185446367)↗
DOE OSTI2025

Quantum AI Based Enhanced Detection of Dementia

Quantum computing has the potential to significantly improve the early detection of Alzheimer's Disease and Related Dementias (ADRD). Quantum-enhanced machine learning can be used to perform an early screening of Alzheimer's disease using brain imaging data based on dataset of MRI scans from both healthy individuals and those diagnosed with Alzheimer's. This study aims to demonstrate the potential of quantum transfer learning to enhance the performance of the classical deep learning model for dementia detection. Using the MRI sagittal images available in the OASIS-2 (64 demented and 72 non-demented subjects between 60 and 96 years), we show how quantum techniques can transform a suboptimal classical model into a more effective solution for dementia detection, highlighting their potential impact on advancing healthcare technology. We begin with a simple classical deep learning model with a significantly smaller number of parameters, which gives suboptimal performance on the problem. Then, we apply different configurations of quantum transfer learning based on the pre-trained weak classifier (Figure 1). We fix the weak classifier's initial convolutional layers at their fixed pre-trained parameters and replace the last set of dense layers with a dressed quantum circuit (DQN), which we train to enhance performance. We performed 4-fold cross-validation for both the classical and the hybrid quantum models and trained them using Pennylane's `default.qubit' simulator and IonQ's Aria-1 simulator (noisy simulation). We showed that with significantly fewer parameters, the quantum transfer learning-based hybrid models showed significant performance enhancement over the base weak classical deep learning model for dementia detection. To classify between a demented and non-demented subject, the accuracy of quantum-based AI methods improved by 6 to 14% compared to classical methods. The sensitivity of the models improved by 4 to 17%. This shows that there are fewer chances of misclassifying demented patients. Figure 2 compares the performance of the hybrid quantum models and their base classical model, and Table 1 summarizes the results. We illustrated that with assistance from quantum machine learning, it is possible to enhance detection for dementia based on brain images. This shows the potential for practical utility of quantum computing in ADRD research.

Bhowmik, Sounak [University of Tennessee, Knoxvill↗
DOE OSTI2025

DONUT: physics-aware machine learning for real-time X-ray nanodiffraction analysis

Coherent X-ray scattering techniques are critical for investigating the fundamental structural properties of materials at the nanoscale. While advancements have made these experiments more accessible, real-time analysis remains a significant bottleneck, often hindered by artifacts and computational demands. In scanning X-ray nanodiffraction microscopy, which is widely used to spatially resolve structural heterogeneities, this challenge is compounded by the convolution of the divergent beam with the sample’s local structure. To address this, we introduce DONUT (Diffraction with Optics for Nanobeam by Unsupervised Training), a physics-aware neural network designed for the rapid and automated analysis of nanobeam diffraction data. By incorporating a differentiable geometric diffraction model directly into its architecture, DONUT learns to predict crystal lattice strain and orientation in real-time. Crucially, this is achieved without reliance on labeled datasets or pre-training, overcoming a fundamental limitation for supervised machine learning in X-ray science. We demonstrate experimentally that DONUT accurately extracts all features within the data over 200 times more efficiently than conventional fitting methods.

Materials science↗
DOE OSTI2025

DONUT: Physics-aware Machine Learning for Real-time X-ray Nanodiffraction Analysis

SF-25-088 Coherent X-ray scattering techniques are critical for investigating the fundamental structural properties of materials at the nanoscale. While advancements have made these experiments more accessible, real-time analysis remains a significant bottleneck, often hindered by artifacts and computational demands. In scanning X-ray nanodiffraction microscopy, which is widely used to spatially resolve structural heterogeneities, this challenge is compounded by the convolution of the divergent beam with the sample’s local structure. To address this, we introduce DONUT (Diffraction with Optics for Nanobeam by Unsupervised Training), a physics-aware neural network designed for the rapid and automated analysis of nanobeam diffraction data. By incorporating a differentiable geometric diffraction model directly into its architecture, DONUT learns to predict crystal lattice strain and orientation in real-time. Crucially, this is achieved without reliance on labeled datasets or pre-training, overcoming a fundamental limitation for supervised machine learning in X-ray science. We demonstrate experimentally that DONUT accurately extracts all features within the data over 200 times more efficiently than conventional fitting methods.

Zhou, Tao [Argonne National Laboratory (ANL), Argo↗
DOE OSTI2025

Benchmarking universal machine learning interatomic potentials for rapid analysis of inelastic neutron scattering data

The accurate calculation of phonons and vibrational spectra remains a significant challenge, requiring highly precise evaluations of interatomic forces. Traditional methods based on the quantum description of the electronic structure, while widely used, are computationally expensive and demand substantial expertise. Emerging universal machine learning interatomic potentials (uMLIPs) offer a transformative alternative by employing pre-trained neural network surrogates to predict interatomic forces directly from atomic coordinates. This approach dramatically reduces computation time and minimizes the need for technical knowledge. In this paper, we produce a phonon database comprising nearly 5000 inorganic crystals to benchmark the performance of several leading uMLIPs. We further assess these models in real-world applications by using them to analyze experimental inelastic neutron scattering data collected on a variety of materials. Through detailed comparisons, we identify the strengths and limitations of these uMLIPs, providing insights into their accuracy and suitability for fast calculations of phonons and related properties, as well as the potential for real-time interpretation of neutron scattering spectra. Our findings highlight how the rapid advancement of AI in science is revolutionizing experimental research and data analysis.

inelastic neutron scattering↗
DOE OSTI2025

Cross-functional transferability in foundation machine learning interatomic potentials

The rapid development of foundation potentials (FPs) in machine learning interatomic potentials demonstrates the possibility for generalizable learning of the universal potential energy surface. The accuracy of FPs can be further improved by bridging the model from lower-fidelity datasets to high-fidelity ones. In this work, we analyze the challenge of this transfer learning (TL) problem within the CHGNet framework. We show that significant energy scale shifts and poor correlations between GGA and r 2 SCAN hinder cross-functional transferability. By benchmarking different TL approaches on the MP-r 2 SCAN dataset, we demonstrate the importance of elemental energy referencing in the TL of FPs. By comparing the scaling law with and without the pre-training on a low-fidelity dataset, we show that significant data efficiency can still be achieved through TL, even with a target dataset of sub-million structures. We highlight the importance of proper TL and multi-fidelity learning in creating next-generation FPs on high-fidelity data.

Huang, Xu [University of California, Berkeley, CA ↗
DOE OSTI2026

Self-supervised and multi-fidelity learning for extended predictive soil spectroscopy

Infrared spectroscopy is a cost-effective, non-destructive, and environmentally benign technology that is increasingly recognized as an important solution for meeting the global demand for soil data. While both near-infrared (NIR) and mid-infrared (MIR) diffuse reflectance spectroscopy enable rapid estimation of soil properties, they present a significant trade-off: NIR offers superior scalability and lower operational costs, whereas MIR provides higher analytical fidelity by capturing fundamental molecular vibrations. In this study, we propose a self-supervised, multi-fidelity learning framework designed to bridge this gap. Our approach leverages large-scale MIR spectral libraries to learn a compact, transferable latent representation, into which NIR spectra are subsequently aligned for downstream prediction. The workflow consists of pretraining a latent model on a large MIR library, adapting the representation using a smaller paired NIR–MIR dataset, and evaluating generalization on an independent external test set. Across a range of chemical and physical soil properties, we found that MIR-derived embeddings improved prediction accuracy relative to baseline models that used raw MIR inputs. Predictions derived from the spectrum conversion (NIR to MIR) task did not match the performance of the original MIR spectra but were similar or superior to predictive performance of NIR-only models, suggesting the unified spectral latent space can effectively leverage the larger and more diverse MIR dataset for prediction of soil properties not well represented in current NIR libraries.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Efficient online quantum circuit learning with no upfront training

Optimization is a promising candidate for studying the utility of variational quantum algorithms (VQAs). However, evaluating cost functions using quantum hardware introduces runtime overheads that limit exploration. Surrogate-based methods can reduce calls to a quantum computer, yet existing approaches require hyperparameter pre-training and have been tested only on small problems. Here, we show that surrogate-based methods can enable successful optimization at scale, without pre-training, by using radial basis function interpolation (RBF) to construct an adaptive, hyperparameter-free surrogate. Using the surrogate as an acquisition function drives hardware queries to the vicinity of the true optima. For 16-qubit random 3-regular Max-Cut instances with the Quantum Approximate Optimization Algorithm (QAOA), our method outperforms state-of-the-art approaches, without considering their upfront training costs. Furthermore, we successfully optimize QAOA circuits for 127-qubit random Ising models on an IBM processor using 10 4 −10 5 measurements. Strong empirical performance demonstrates the promise of automated surrogate-based learning for large-scale VQA applications.

97 MATHEMATICS AND COMPUTING↗