Search NASA⌕ Search

SEARCH · Search NASA

Results for “interpretable 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 181 records · Page 10
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

LATTE: open-source, high-performance traveltime computation, tomography and source location in acoustic and elastic media

Traveltime-based tomography and source location are fundamental approaches for imaging subsurface structures and understanding the spatiotemporal distribution of seismicity from local to global scales. We present an open-source, high-performance framework integrating eikonal equation solvers and adjoint-state theory for traveltime computation, velocity tomography, source location and joint tomography-location in 2-D/3-D acoustic and elastic media. We introduce novel regularization schemes based on total generalized p-variation, structural similarity and multitask machine learning to enhance the fidelity and interpretability of inverted models and source locations. Key features of our implementation also include the ability to leverage both absolute-difference and double-difference traveltime misfits for high-fidelity velocity tomography and source parameter estimation; support for traveltime computation and inversion in diverse 2-D/3-D scenarios with arbitrary source and receiver distributions; and a perturbation-based optimal step-size estimation method to reduce computational costs. In addition, our implementation employs shared-memory and distributed-memory parallelization to provide an efficient solution for traveltime computation, tomography, and source location. In conclusion, we validate the efficacy and accuracy of our approach through multiple synthetic data examples.

58 GEOSCIENCES↗
DOE OSTI2024

Applications of a Rayleigh-Taylor model to direct-drive laser fusion

Here, this paper presents a simple physics-based model for the interpretation of key metrics in laser direct drive. The only input parameters required are target scale, in-flight aspect ratio, and beam-to-target radius, and the importance of each has been quantified with a tailored set of cryogenic implosion experiments. These analyses lead to compact and accurate predictions of the fusion yield and areal density as a function of hydrodynamic stability, and suggest new ways to take advantage of direct drive. To provide examples, we will discuss how the inferred mix width behaves relative to theory, then show how it could be exploited to perform a direct drive implosion with a Lawson metric or Χ no α of 0.24 ± 0.02—using a novel parameter space at high velocities and beam radii on the OMEGA laser—that projects to ignition at a laser energy ≤2.0 MJ.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2026

buhito

buhito is a Python library for graph analysis and machine learning. Graphs can represent networks with objects as nodes and their relationships as edges. buhito focuses on graphlet methods that study graphs through enumerating their component subgraphs to enable interpretable and fast models of complex systems. The package provides tools for different algorithmic designs for computing, analyzing, and applying graphlets to research problems such as machine learning, data compression, and anomaly detection in graph-structured data. A central feature is performing decomposition data analysis on graphs for machine learning models. Implemented in Python and built upon open-source scientific libraries such as NetworkX, NumPy, and SciPy, buhito provides high-performance methods for researchers exploring the mathematical and computational foundations of graphlet analysis applicable to systems of different sizes.

Pimonova, Yulia↗
DOE OSTI2025

Ranking Biological Features in Soil-Based Microbial Multi-Omics Data with Integration Modeling

Distinguishing the most important features (e.g. proteins, metabolites, etc.) per group (e.g. control and treatment) is a critical challenge in feature-rich multi-omics experiments, especially in soil data. Traditional feature identification and ranking approaches, such as differential expression, are based on single omics and thus not directly translatable to multi-omics experiments. Here, 5 multi-omics integration models (DIABLO, JACA, MOFA, MultiMLP, and SLIDE) that were not explicitly built for soil data applications were tested using a soil-based multi-omics experiment. The data were obtained from an experimental setup of an autoclaved soil system inoculated with 8 bacteria and using chitin as the carbon source and including samples collected at 0- (control), 4-, 8-, and 12-weeks post-inoculation. The omics data included metaproteomics, 16S rRNA sequencing, and LC-MS/MS metabolomics (in positive and negative mode). Each multi-omics integration model was implemented, and top features were compared to differential univariate statistics per omic type, demonstrating that integration approaches cut the potential number of top features from 2957 identified by differential statistics to 13-224 (a 99.6% to 92.4% reduction). Interestingly, most top features across integration models were not shared; though, scaling and averaging ranks across models shared similar patterns. This work highlights the usefulness of multi-omics integration models in soil-based microbial studies and the power of using multiple integration models together to interpret results.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Radiation GRMHD Models of Accretion onto Stellar-mass Black Holes. I. Survey of Eddington Ratios

We summarize results from a survey of radiation-dominated black hole accretion flows across a wide range of mass accretion rates, as well as two values of black hole spin and initial magnetic field geometry. These models apply an algorithm targeting direct solutions to the radiation transport equation in full general relativity and have been enabled by access to modern exascale computing systems. Super-Eddington accretion flows form geometrically thick radiation-pressure-supported disks that drive powerful equatorial outflows. A narrow funnel-shaped photosphere in the inner region results in very low radiative efficiencies in this regime. The structure of near- and sub-Eddington accretion depends on whether there is net vertical magnetic flux at the midplane of the disk. With net flux, the disk forms a thin, dense layer at the midplane surrounded by a magnetically dominated corona, whereas without net flux, the disk remains magnetically dominated everywhere. Although none of our models achieve the magnetically arrested disk regime, those with net vertical flux and a rapidly spinning black hole still produce powerful relativistic jets. Our calculations adopt simple opacity models (with scalings appropriate to stellar-mass black hole accretion). We discuss the application of our results to observations of X-ray binaries and ultraluminous X-ray sources such as Cyg X-3 and SS433. We also speculate on the application of our super-Eddington models to the interpretation of little red dots recently discovered by JWST.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2025

Edge ML for CAN bus intrusion detection in AVs

Autonomous Vehicles (AVs) are revolutionizing transportation, but their reliance on interconnected cyber-physical systems exposes them to unprecedented cybersecurity risks. This study addresses the critical challenge of detecting real-time cyber intrusions in self-driving vehicles by leveraging a dataset from the Udacity self-driving car project. We simulate four high-impact attack vectors, Denial of Service (DoS), spoofing, replay, and fuzzy attacks, by injecting noise into spatial features (e.g., bounding box coordinates) to replicate adversarial scenarios. We develop and evaluate two lightweight neural network architectures (NN-1 and NN-2) alongside a logistic regression baseline (LG-1) for intrusion detection. The models achieve exceptional performance, with NN-2 attaining an AUC score of 93.15% and 93.15% accuracy, demonstrating their suitability for edge deployment in AV environments. Through explainable AI techniques, we uncover unique forensic fingerprints of each attack type, such as spatial corruption in fuzzy attacks and temporal anomalies in replay attacks, offering actionable insights for feature engineering and proactive defense. Visual analytics, including confusion matrices, ROC curves, and feature importance plots, validate the models' robustness and interpretability. This research sets a new benchmark for AV cybersecurity, delivering a scalable, field-ready toolkit for Original Equipment Manufacturers (OEMs) and policymakers. By aligning intrusion fingerprints with SAE J3061 automotive security standards, we provide a pathway for integrating machine learning into safety-critical AV systems. Our findings underscore the urgent need for security-by-design AI, ensuring that AVs not only drive autonomously but also defend autonomously. This work bridges the gap between theoretical cybersecurity and life-preserving engineering, offering a leap toward safer, more secure autonomous transportation.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Search for dark matter produced in association with a pair of bottom quarks in proton-proton collisions at $\sqrt{s}$ = 13 TeV

A search for dark matter (DM) particles produced in association with bottom quarks is presented. The analysis uses proton-proton collision data at a center-of-mass energy of $\sqrt{s}$ = 13 TeV, corresponding to an integrated luminosity of 138 fb -1 . The search is performed in a final state with large missing transverse momentum and a pair of jets originating from bottom quarks. No significant excess of data is observed with respect to the standard model expectation. Results are interpreted in the context of a type-II two-Higgs-doublet model with an additional light pseudoscalar (2HDM+a). An upper limit is set on the mass of the lighter pseudoscalar, probing masses up to 260 GeV at 95% confidence level. Sensitivity to the parameter space with the ratio of the vacuum expectation values of the two Higgs doublets, tan β, greater than 15 is achieved, capitalizing on the enhancement of couplings between pseudoscalars and bottom quarks with high tan β.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

Towards verifiable cancer digital twins: tissue level modeling protocol for precision medicine

Cancer exhibits substantial heterogeneity, manifesting as distinct morphological and molecular variations across tumors, which frequently undermines the efficacy of conventional oncological treatments. Developments in multiomics and sequencing technologies have paved the way for unraveling this heterogeneity. Nevertheless, the complexity of the data gathered from these methods cannot be fully interpreted through multimodal data analysis alone. Mathematical modeling plays a crucial role in delineating the underlying mechanisms to explain sources of heterogeneity using patient-specific data. Intra-tumoral diversity necessitates the development of precision oncology therapies utilizing multiphysics, multiscale mathematical models for cancer. This review discusses recent advancements in computational methodologies for precision oncology, highlighting the potential of cancer digital twins to enhance patient-specific decision-making in clinical settings. We review computational efforts in building patient-informed cellular and tissue-level models for cancer and propose a computational framework that utilizes agent-based modeling as an effective conduit to integrate cancer systems models that encode signaling at the cellular scale with digital twin models that predict tissue-level response in a tumor microenvironment customized to patient information. Furthermore, we discuss machine learning approaches to building surrogates for these complex mathematical models. These surrogates can potentially be used to conduct sensitivity analysis, verification, validation, and uncertainty quantification, which is especially important for tumor studies due to their dynamic nature.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2025

Airport Delay Prediction with Temporal Fusion Transformers

Since flight delay hurts passengers, airlines, and airports, its prediction becomes crucial for the decision-making of all stakeholders in the aviation industry and thus has been attempted by various previous research. However, previous delay predictions are often categorical and at a highly aggregated level. To improve that, this study proposes to apply the novel Temporal Fusion Transformer model and predict numerical airport arrival delays at quarter hour level for U.S. top 30 airports. Inputs to our model include airport demand and capacity forecasts, historic airport operation efficiency information, airport wind and visibility conditions, as well as en-route weather and traffic conditions. The results show that our model achieves satisfactory performance measured by small prediction errors on the test set. In addition, the interpretability analysis of the model outputs identifies the important input factors for delay prediction.

Liu, Ke [University of California Berkeley]↗
DOE OSTI2025

The Role of Water Volume Fraction on Water Adsorption in Anion Exchange Membranes

Water absorption plays a key role in the performance of polymeric anion exchange membranes. It influences important properties such as ionic conductivity and mechanical strength and alters their performance as solid electrolytes in hydrogen electrochemical devices for energy conversion. However, computational approaches that address the relationship between the polymer design and the absorption process are scarce. In this work, we introduce a simple thermodynamic model to predict the water absorption isotherms of polyelectrolyte membranes in contact with a water vapor reservoir that incorporates the specific chemical design of the polymers. The model accurately predicts the water content and macrostructural properties of polynorbornene membranes as a function of the water activity and successfully captures the effect of various polymer design parameters. The energy of pairwise attractive interactions predicted by our model provides a means to interpret the absorption process at the molecular level. The model also reveals the most significant favorable and unfavorable contributions to the free energy and indicates that their balance is solely governed by the water volume fraction, regardless of the polymer design. This universal behavior leads to important implications in the search for better ion exchange membranes.

Zaldivar, Gervasio [New York University, Brooklyn,↗
DOE OSTI2024

A novel method may reveal bulk metallic glass compressive ductility trends in high data rate nanoindentation

Recent methods allow novel amorphous alloy compositions to be rapidly manufactured at small scale; however, obtaining materials properties such as compressive ductility from these smaller specimens has remained a challenge. Here, we suggest a potential high-throughput nanoindentation method that may be able to rapidly characterize the relative compressive ductility between these alloys based on their serration characteristics. The properties of emergent serrations, when interpreted in a simple micromechanical stress relaxation model, may order these materials by their compressive plastic strain to failure. These results are consistent with the ordering obtained from compressed specimens as well as with model simulations, suggesting that this model may be broadly useful for interpreting compressive ductility from nanoindentation serrations. After it is validated on more materials, this new method will match the rapid pace of amorphous alloy development, thus allowing metallic glass properties to be fine-tuned for each application prior to scale prototyping.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Model-independent search for pair production of new bosons decaying into muons in proton-proton collisions at $\sqrt{s}$ = 13 TeV

The results of a model-independent search for the pair production of new bosons within a mass range of 0.21 < m < 60 GeV, are presented. This study utilizes events with a four-muon final state. We use two data sets, comprising 41.5 fb −1 and 59.7 fb −1 of proton-proton collisions at $\sqrt{s}$ = 13 TeV, recorded in 2017 and 2018 by the CMS experiment at the CERN LHC. The study of the 2018 data set includes a search for displaced signatures of a new boson within the proper decay length range of 0 < cτ < 100 mm. Our results are combined with a previous CMS result, based on 35.9 fb −1 of proton-proton collisions at $\sqrt{s}$ = 13 TeV collected in 2016. No significant deviation from the expected background is observed. Results are presented in terms of a model-independent upper limit on the product of cross section, branching fraction, and acceptance. The findings are interpreted across various benchmark models, such as an axion-like particle model, a vector portal model, the next-to-minimal supersymmetric standard model, and a dark supersymmetric scenario, including those predicting a non-negligible proper decay length of the new boson. In all considered scenarios, substantial portions of the parameter space are excluded, expanding upon prior results.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

Accelerating the design of lattice structures using machine learning

Lattices remain an attractive class of structures due to their design versatility; however, rapidly designing lattice structures with tailored or optimal mechanical properties remains a significant challenge. With each added design variable, the design space quickly becomes intractable. To address this challenge, research efforts have sought to combine computational approaches with machine learning (ML)-based approaches to reduce the computational cost of the design process and accelerate mechanical design. While these efforts have made substantial progress, significant challenges remain in (1) building and interpreting the ML-based surrogate models and (2) iteratively and efficiently curating training datasets for optimization tasks. Here, we address the first challenge by combining ML-based surrogate modeling and Shapley additive explanation (SHAP) analysis to interpret the impact of each design variable. We find that our ML-based surrogate models achieve excellent prediction capabilities (R 2 > 0.95) and SHAP values aid in uncovering design variables influencing performance. We address the second challenge by utilizing active learning-based methods, such as Bayesian optimization, to explore the design space and report a 5 × reduction in simulations relative to grid-based search. Collectively, these results underscore the value of building intelligent design systems that leverage ML-based methods for uncovering key design variables and accelerating design.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Unrolled Video Super-Resolution Network with Autoregressive Prior for the Case of Known Motion

Real-time detection and classification of distant objects is necessary for many national security applications. However, when objects are far from the sensor, they occupy only a small number of pixels in the captured video, limiting the amount of visual detail available for recognition. State-of-the-art classification methods typically rely on high-resolution (HR) video streams to capture characteristic object features, but obtaining such detail is challenging for distant objects that occupy only a few pixels. This motivates the development of video super-resolution (VSR) methods that enhance object classification by recovering fine details from low-pixel representations. Current VSR methods rely either on model-based optimization, which is interpretable but computationally expensive, or on learning-based approaches, which are efficient and high-performing but often lack flexibility and interpretability. In this report, we propose an end-to-end trainable unrolled VSR network, UVSRNet, which super-resolves each frame in a video by exploiting sub-pixel motion between neighboring low-resolution (LR) frames as well as incorporating high-frequency detail from previously super-resolved frames. In particular, by unrolling a plug-and-play (PnP) half-quadratic splitting (HQS) algorithm, we leverage a model-based data-fitting module alongside a learning-based autoregressive prior module. This combination yields a method that maintains the flexibility and interpretability of model-based methods while achieving the performance advantages of learning-based methods.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗
DOE OSTI2024

Combined search for electroweak production of winos, binos, higgsinos, and sleptons in proton-proton collisions at s = 13 TeV

A combination of the results of several searches for the electroweak production of the supersymmetric partners of standard model bosons, and of charged leptons, is presented. All searches use proton-proton collision data at s = 13 TeV recorded with the CMS detector at the LHC in 2016–2018. The analyzed data correspond to an integrated luminosity of up to 137 fb − 1 . The results are interpreted in terms of simplified models of supersymmetry. Two new interpretations are added with this combination: a model spectrum with the bino as the lightest supersymmetric particle together with mass-degenerate Higgsinos decaying to the bino and a standard model boson, and the compressed-spectrum region of a previously studied model of slepton pair production. Improved analysis techniques are employed to optimize sensitivity for the compressed spectra in the wino and slepton pair production models. The results are consistent with expectations from the standard model. The combination provides a more comprehensive coverage of the model parameter space than the individual searches, extending the exclusion by up to 125 GeV, and also targets some of the intermediate gaps in the mass coverage. © 2024 CERN, for the CMS Collaboration 2024 CERN

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2026

Search for HH → bbτ⁺τ⁻ Using Run 3 Scouting Data Analyze b-tagging and tau-tagging Performance with Unified Particle Transformer

B-tagging and tau-tagging performances play an important role in the search for the rare event HH → bbτ⁺τ⁻. A transformer-based neural network, Unified Particle Transformer, is applied for both tagging tasks, and Run 3 proton–proton collision scouting data at center-of-mass energy of 13.6 TeV is used. The scouting data stream accepts events at a much higher rate compared to traditional triggers, but stores only the objects reconstructed in the trigger, no low-level detector information. Therefore, existing taggers trained for the offline event reconstruction cannot be used. Analysis of the SoftMax plots, ROC/AUC curves, confusion matrix, accuracy and losses are used to evaluate model performance. Specifically, the tagging efficiency of the signal and misidentification probability across multiple background processes are compared for varying working points. Different training samples with distinct distributions of jet flavors are utilized and related model performances are analyzed. Interpretability methods, such as Integrated Gradients, may further be applied to study the input features’ influence on the model’s decisions, providing insights into potential improvements.

Chen, Blair [Purdue U., West Lafayette; Fermilab]↗
DOE OSTI2025

Model ages of three uranium metal CRMs and implications for radiochronometry data interpretation

In this study, we report 230 Th/ 234 U and 231 Pa/ 235 U model ages in three uranium metal certified reference materials with distinct production histories and isotopic compositions ranging from depleted to highly enriched (CRMs 112-A, 115, and 116-A). Here, our results provide the first model age values for these materials using multiple radiochronometers, which can be used as reference values for quality control of radiochronometry measurements performed during nuclear forensic investigations. Furthermore, we evaluate the model ages in the context of available production records for the materials, which provides insight into the processes that may impact radiochronometry systems during uranium metal production.

38 RADIATION CHEMISTRY, RADIOCHEMISTRY, AND NUCLEA↗
DOE OSTI2024

Evaluating the Effectiveness of Retrieval-Augmented Large Language Models in Scientific Document Reasoning

Despite the dramatic progress in Large Language Model (LLM) development, LLMs often provide seemingly plausible but not factual information, often referred as hallucinations. Retrieval-augmented LLMs provide a non-parametric approach to solve these issues by retrieving relevant information from external data sources and augment the training process. These models helps to trace evidence from an externally provided knowledge base allowing the model predictions to be better interpreted and verified. In this work, we critically evaluate these models in their ability to perform in scientific document reasoning tasks. To this end, we tuned multiple such model variants with science-focused instructions and evaluated them on a scientific document reasoning benchmark for the usefulness of the retrieved document passages. Our findings suggest that models justify predictions in science tasks with fabricated evidence and leveraging scientific corpus as pretraining data does not alleviate the risk of evidence fabrication.

• Artificial intelligence (AI) / machine learning ↗