Search NASA⌕ Search

SEARCH · Search NASA

Results for “Interpretability”

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 91 records · Page 5
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

MoRE-Brain: Routed Mixture of Experts for Interpretable and Generalizable Cross-Subject fMRI Visual Decoding

Decoding visual experiences from fMRI offers a powerful avenue to understand human perception and develop advanced brain-computer interfaces. However, current progress often prioritizes maximizing reconstruction fidelity while overlooking interpretability, an essential aspect for deriving neuroscientific insight. To address this gap, we propose MoRE-Brain, a neuro-inspired framework designed for high-fidelity, adaptable, and interpretable visual reconstruction. MoRE-Brain uniquely employs a hierarchical Mixture-of-Experts architecture where distinct experts process fMRI signals from functionally related voxel groups, mimicking specialized brain networks. The experts are first trained to encode fMRI into the frozen CLIP space. A finetuned diffusion model then synthesizes images, guided by expert outputs through a novel dual-stage routing mechanism that dynamically weighs expert contributions across the diffusion process. MoRE-Brain offers three main advancements: First, it introduces a novel Mixture-of-Experts architecture grounded in brain network principles for neuro-decoding. Second, it achieves efficient cross-subject generalization by sharing core expert networks while adapting only subject-specific routers. Third, it provides enhanced mechanistic insight, as the explicit routing reveals precisely how different modeled brain regions shape the semantic and spatial attributes of the reconstructed image. Extensive experiments validate MoRE-Brain’s high reconstruction fidelity, with bottleneck analyses further demonstrating its effective utilization of fMRI signals, distinguishing genuine neural decoding from over-reliance on generative priors. Consequently, MoRE-Brain marks a substantial advance towards more generalizable and interpretable fMRI-based visual decoding.

Wei, Yuxiang [Georgia Institute of Technology]↗
DOE OSTI2025

SPARTA: A flux adjustment methodology to interpret complex experiments

For the accurate determination of reactivity from a detector count rate, correction of spatial effects is of prime importance. This spatial correction is often provided using simulation methodologies, but this may introduce a bias if the result of the experiment is also used as input data for the simulation. Here, this work presents a flux adjustment methodology able to infer experimental reactivity and correction of spatial effects without the need for a simulation. It can process the signal from a complex experiment such as a heat balance measurement in the TREAT reactor, where control rods are continuously adjusted to maintain a constant power. In the present work, this methodology successfully computed the reactivity and the local spatial variation of the flux of a generated signal. It also proved to be robust against noise and errors on kinetic parameters and provides a credible interpretation of a heat balance experiment in TREAT. Efficiency of flux adjustment methods for complex experiment enable a better experiment interpretation less reliant on nuclear data evaluation.

73 - NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2024

Analyzing the impact of design factors on solar module thermomechanical durability using interpretable machine learning techniques

Solar modules in utility-scale systems are expected to maintain decades of lifetime to rival conventional energy sources. However, cyclic thermomechanical loading often degrades their long-term performance, highlighting the importance of effective design to mitigate thermal expansion mismatches between module materials. Given the complex composition of solar modules, isolating the impact of individual components on overall durability remains a challenging task. In this work, we analyze a comprehensive data set that comprises bill-of-materials (BOM) and thermal cycling power loss from 251 distinct module designs to identify the predominant design factors and their impacts on the thermomechanical durability of modules. The methodology of our analysis combines machine learning modeling (random forest) and Shapley additive explanation (SHAP) to correlate design factors with power loss and interpret the model’s decision-making. The interpretation reveals that silicon type (monocrystalline or polycrystalline), encapsulant thickness, busbar numbers, and wafer thickness predominantly influence the degradation. With lower power loss of around 0.6% on average in the SHAP analysis, monocrystalline cells present better durability than polycrystalline cells. This finding is further substantiated by statistical testing on our raw data set. The SHAP analysis also demonstrates that while thicker encapsulants lead to reduced power loss, further increasing their thickness over around 0.6 to 0.7 mm does not yield additional benefits, particularly for the front side one. In addition, other important BOM features such as the number of busbars are analyzed. This study provides a blueprint for utilizing explainable machine learning techniques in a complex material system and can potentially guide future research on optimizing the design of solar modules.

14 SOLAR ENERGY↗
DOE OSTI2026

Interpretable and flexible non-intrusive reduced-order models using reproducing kernel Hilbert spaces

This paper develops an interpretable, non-intrusive reduced-order modeling technique using regularized kernel interpolation. Existing non-intrusive approaches approximate the dynamics of a reduced-order model (ROM) by solving a data-driven least-squares regression problem for low-dimensional matrix operators. Our approach instead leverages regularized kernel interpolation, which yields an optimal approximation of the ROM dynamics from a user-defined reproducing kernel Hilbert space. We show that our kernel-based approach can produce interpretable ROMs whose structure mirrors full-order model structure by embedding judiciously chosen feature maps into the kernel. The approach is flexible and allows a combination of informed structure through feature maps and closure terms via more general nonlinear terms in the kernel. We also derive a computable a posteriori error bound that combines standard error estimates for intrusive projection-based ROMs and kernel interpolants. In conclusion, the approach is demonstrated in several numerical experiments that include comparisons to operator inference using both proper orthogonal decomposition and quadratic manifold dimension reduction.

Data-driven model reduction↗
DOE OSTI2026

Learning interpretable surface elasticity properties from bulk properties via neural network equation learners

Surface elasticity is central to understanding the mechanics and stability of surfaces and interfaces. It is characterized by quantities such as surface tension, residual surface stress, and surface stiffness. However their analytical expressions are typically difficult to derive from atomistic data, and depend strongly on modeling choices. This work presents a neural network-based equation learner which combines customized activation functions and connection-based pruning to discover parsimonious, closed-form equations for surface elasticity from atomistic simulations. Applying the method to seven face-centered cubic (FCC) metals, our equation learner uncovers interpretable equations that describe both low-Miller index and high-Miller index surface properties, capturing long-tail property distributions accurately. The discovered expressions are decoupled into two components: a universal, geometry-driven orientation function, and material-specific baseline coefficients. We find that lower-order properties such as surface tension are fundamentally geometry dependent, while higher-order properties such as surface stress and elasticity show more complex geometry and material dependence. We also relate material dependent coefficients to bulk properties, forming a clear map from bulk material properties to surface elasticity. Overall, this approach demonstrates that interpretable neurosymbolic machine learning can bridge the gap between atomistic simulations and physical laws, enabling the discovery of generalizable structure–property relationships for materials science phenomena such as surface elasticity.

Equation learning↗
DOE OSTI2025

Self-Supervised and Interpretable Anomaly Detection Using Network Transformers

Machine learning and deep neural networks (DNNs) have been proposed as a tool to identify anomalies in computer network communications. However, due the obfuscated nature of off-the-shelf machine learning models, their output often does not provide enough information to isolate the source of the anomaly to take corrective measures. In this article, we introduce the network transformer (NeT), a DNN model for anomaly detection that incorporates the graph structure of the communication network in order to improve interpretability. Further, the presented approach has the following advantages: first, enhanced interpretability by incorporating the graph structure of computer networks; second, provides a hierarchical set of features that enables analysis at different levels of granularity; second, self-supervised training that does not require labeled data. The NeT model was evaluated on a set of anomalous scenarios executed in a real industrial control system. The presented approach successfully identified the anomalies, the devices affected, and the specific connections causing the anomalies, providing a data-driven hierarchical approach to analyze the behavior of a cyber network.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Contrasting Carbon–Water–Energy Dynamics in Perennial and Annual Bioenergy Agroecosystems Using Eddy Covariance and Interpretable Machine Learning

Understanding how agroecosystems respond to environmental variability is fundamental to predicting productivity and sustainability under a changing climate. We analyzed 55 site-years of high-frequency eddy covariance observations from five agroecosystems—two perennial grasses (miscanthus and switchgrass), two annual rotation systems (maize–soybean and sorghum–soybean), and a restored native prairie—to examine ecosystem-scale carbon, water, and energy fluxes. Using an interpretable machine-learning framework with regression tree ensembles, Shapley Additive Explanations, and Accumulated Local Effects, we quantified how environmental and temporal factors regulate gross primary productivity (GPP), evapotranspiration (ET), water-use efficiency, and the Bowen ratio. Perennials exhibited stronger physiological buffering and maintained fluxes across a broader range of temperature and moisture conditions, reflecting deeper rooting and persistent canopy cover. Annuals, in contrast, showed greater short-term variability and stronger coupling to atmospheric demand, with GPP and ET declining rapidly under low humidity or soil moisture. Differences in temperature sensitivity of Bowen ratio further revealed that perennials sustained proportionally greater sensible heat flux under cool conditions, whereas annuals exhibited constrained energy exchange when evaporative demand was low. Together, these results demonstrate that crop life cycle and canopy structure are fundamental determinants of ecosystem-scale carbon–water–energy coupling. By integrating long-term flux observations with interpretable machine learning, this study identifies the environmental drivers that shape agroecosystem function and highlights how conversion from annual to perennial feedstocks can enhance climatic resilience and alter land–atmosphere energy feedbacks. These findings provide a data-driven basis for improving crop and Earth-system models and for guiding bioenergy landscape design under future climate scenarios.

Accumulated Local Effects↗
DOE OSTI2024

Interpretations of the ATLAS measurements of Higgs boson production and decay rates and differential cross-sections in pp collisions at $\sqrt{s}$ = 13 TeV

Measurements of the Higgs boson production times decay rates and differential cross-sections have recently been performed by the ATLAS experiment in several decay channels using up to 139 fb -1 of proton-proton collision data at $\sqrt{s}$ = 13 TeV recorded at the Large Hadron Collider. This paper presents multiple interpretations of these Higgs boson measurements. Measurements of production-mode cross-sections, simplified template cross-sections and fiducial differential cross-sections in different decay channels are reparameterised in terms of the impact of Standard Model effective field theory operators, and constraints are reported on the corresponding Wilson coefficients. Production and decay rate measurements are interpreted in UV-complete extensions of the Standard Model, namely the two-Higgs-doublet model (2HDM) near the alignment limit and the Minimal Supersymmetric Standard Model (MSSM) for various MSSM benchmark scenarios. The constraints on the 2HDM parameters (cos(β - α), tanβ) and the MSSM parameters (m A , tanβ) are complementary to those obtained from direct searches for additional Higgs bosons.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

Interpretable, extensible linear and symbolic regression models for charge density prediction using a hierarchy of many-body correlation descriptors

Here, density functional theory (DFT) is routinely used to make electronic structure predictions for high-throughput screening of materials and molecules for technologically relevant areas, like the identification of better catalysts, electronic materials, and drug discovery. However, the DFT formalism is limited by (a) its poor (quadratic-to-quartic) scaling, and (b) the need to perform repeated eigenvalue computations of the electronic Hamiltonian as part of its self-consistent field (SCF) iteration procedure to obtain the converged ground state electron density, ρ (r). Approaches that directly predict ρ (r) of a structure with high accuracy can accelerate conventional SCF calculations and can also be used in linearly scaling methods such as orbital-free DFT. To this end, we present a procedure to predict the ground state electron density of molecular and periodic three-dimensional systems directly from the atomic structure with a particular emphasis on physical interpretability. In our framework, ρ (r) is modeled using many-body correlation descriptors that accurately capture the effects of local atomic arrangements in the neighborhood of a grid point. Our use of a linear regression scheme to fit to charge density data enables transparent analysis of the relative contributions of various types of local atomic correlations. By systematically including increasingly complex correlations, our model is shown to accurately predict ρ (r) for a variety of chemically and electronically diverse systems — amorphous Ge, Al(001) slab, crystalline Ga 2 O 3 , molecular benzene, and polyethylene. We then demonstrate a symbolic regression-based protocol to construct easily computable, interpretable features from lower-order correlations that significantly improves our electron density predictions with effectively no increase in the computational cost.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2026

Interpretation of high-harmonic fast-wave propagation in the scrape-off layer of NSTX-U as a geometrically bounded-waveguide mode

This paper revisits the onset of high-harmonic fast-wave (HHFW) propagation in the scrape-off layer (SOL) plasma of the NSTX/NSTX-U spherical tokamak, motivated by past HHFW heating and current drive experiments and modelling. Previously, the fast-wave propagation in the SOL was correlated with the opening (suppression) of the fast-wave right-hand cutoff layer in front of the antenna. In this work, the SOL propagation is interpreted as a geometric waveguide mode guided by the SOL geometry, whose radial width can be comparable to the wave’s perpendicular wavelength in the HHFW regime. A two-dimensional circular model is first employed using full-wave solvers to characterise the poloidal eigenmode structures and to clarify their relationship to annulus resonance. By progressively adding a tokamak-like magnetic-field configuration, starting from a uniform axial field, the effects of magnetic-field gradients and pitch on the SOL waveguide mode are characterised. The poloidal mode numbers supported by the SOL plasma agree well with analytic estimates, indicating that an anisotropic plasma in a bounded geometry selectively supports and amplifies the resonant poloidal mode number. Additionally, two-dimensional axisymmetric NSTX-U simulations demonstrate that the SOL eigenmode features identified in the circular model persist in the experimentally relevant configurations. A control approach based on lengthening the wave perpendicular wavelength relative to the SOL width, including a higher magnetic (B)-field operation, is discussed. The analysis here shows the key role of a bounded geometry in interpreting HHFW eigenmode coupling and propagation in the NSTX-U SOL plasma.

plasma heating↗
DOE OSTI2026

Deciphering the Solvation Structure of Aqueous ZnCl 2 Solutions from X-ray Absorption Spectra Using the Interpretable Graph Neural Network

Machine learning (ML) provides powerful pathways for predicting spectroscopic observables from atomic structures, but its broader impact depends on making model predictions interpretable in terms of physical and chemical principles. Here, we introduce a physics-guided graph neural network (GNN) model that predicts Zn K-edge X-ray spectroscopy (XAS) spectra of aqueous ZnCl 2 solutions. Training data are generated from ab initio XAS calculations on molecular dynamics snapshots obtained using a machine learning interatomic potential. The GNN reproduces experimental spectra across concentrations from dilute (<0.1 m) to highly concentrated (30 m, “water-in-salt”) regimes and scales efficiently to large, disordered liquid systems beyond the reach of conventional ab initio approaches. Gradient-based attribution analysis reveals that the model learns physically meaningful structure-spectrum relationships. Ligand-specific attributions reflect orbital hybridization patterns and the origin of the excitations derived from the density functional theory. Bond-length attributions recover spectral shifts consistent with multiple-scattering theory. Finally, this work bridges data-driven prediction with electronic-structure theory, establishing a general paradigm for interpretable ML that links atomic structure, electronic structure, and spectroscopic observables.

25 ENERGY STORAGE↗
DOE OSTI2025

Learning Physically Interpretable Atmospheric Models From Data With WSINDy

The multiscale and turbulent nature of Earth's atmosphere has historically rendered accurate weather modeling a hard problem. Recently, there has been an explosion of interest surrounding data-driven approaches to weather modeling, which in many cases show improved forecasting accuracy and computational efficiency when compared to traditional methods. However, many of the current data-driven approaches employ highly parameterized neural networks, often resulting in uninterpretable models and limited gains in scientific understanding. In this work, we address the interpretability problem by explicitly discovering partial differential equations governing atmospheric phenomena, identifying symbolic mathematical models with direct physical interpretations. The purpose of this paper is to demonstrate that, in particular, the weak-form sparse identification of nonlinear dynamics (WSINDy) algorithm can learn effective atmospheric models from both simulated and assimilated data. Our approach adapts the standard WSINDy algorithm to work with high-dimensional fluid data of arbitrary spatial dimension.

58 GEOSCIENCES↗
DOE OSTI2025

Evaluating the factors influencing accuracy, interpretability, and reproducibility in the use of machine learning classifiers in biology to enable standardization

The complexity and variability of biological data has promoted the increased use of machine learning methods to understand processes and predict outcomes. These same features complicate reliable, reproducible, interpretable, and responsible use of such methods, resulting in questionable relevance of the derived. outcomes. Here we systematically explore challenges associated with applying machine learning to predict and understand biological processes using a well- characterized in vitro experimental system. We evaluated factors that vary while applying machine learning classifers: (1) type of biochemical signature (transcripts vs. proteins), (2) data curation methods (pre- and post-processing), and (3) choice of machine learning classifier. Using accuracy, generalizability, interpretability, and reproducibility as metrics, we found that the above factors significantly mod- ulate outcomes even within a simple model system. Our results caution against the unregulated use of machine learning methods in the biological sciences, and strongly advocate the need for data standards and validation tool-kits for such studies.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Informed total-error-minimizing priors: Interpretable cosmological parameter constraints despite complex nuisance effects

While Bayesian inference techniques are standard in cosmological analyses, it is common to interpret resulting parameter constraints with a frequentist intuition. This intuition can fail, for example, when marginalizing high-dimensional parameter spaces onto subsets of parameters, because of what has come to be known as projection effects or prior volume effects. We present the method of informed total-error-minimizing (ITEM) priors to address this problem. An ITEM prior is a prior distribution on a set of nuisance parameters, such as those describing astrophysical or calibration systematics, intended to enforce the validity of a frequentist interpretation of the posterior constraints derived for a set of target parameters (e.g., cosmological parameters). Our method works as follows. For a set of plausible nuisance realizations, we generate target parameter posteriors using several different candidate priors for the nuisance parameters. We reject candidate priors that do not accomplish the minimum requirements of bias (of point estimates) and coverage (of confidence regions among a set of noisy realizations of the data) for the target parameters on one or more of the plausible nuisance realizations. Of the priors that survive this cut, we select the ITEM prior as the one that minimizes the total error of the marginalized posteriors of the target parameters. As a proof of concept, we applied our method to the density split statistics measured in Dark Energy Survey Year 1 data. We demonstrate that the ITEM priors substantially reduce prior volume effects that otherwise arise and that they allow for sharpened yet robust constraints on the parameters of interest.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2025

Interpreting scattered neutron spectra on omega cryogenic implosions with measured backgrounds and higher-order scattering

Measurements of the areal density (⁠ρR⁠) of inertially confined implosions are critical to evaluate their performance. On OMEGA, ρR is inferred from measurements of the scattered neutron spectrum via neutron-time-of-flight (nTOF) and magnetic recoil spectrometer detectors. The nTOF measurements, in particular, have the ability to measure a wide range of neutron energies and thus scattering angles, with high precision. However, the neutrons that are backscattered into the detector must transit through the entire dense fuel assembly, and as a result are subject to rescattering effects, which have been heretofore neglected in the interpretation of OMEGA neutron spectra. At the backscatter edge, neglecting rescattering on OMEGA can lead to a ~10 to 20% reduction of the apparent areal density and therefore must be included in the analysis. The low ρR s on OMEGA also mean that backgrounds from non-target physics effects, such as scattering in the target chamber or slow scintillator decays, can significantly alter the measured signal. Here, we discuss how the Monte Carlo neutron spectrum postprocessor IRIS is used to include second-order scattering effects in the interpretation of the scattered neutron spectrum in OMEGA implosion experiments, and how dedicated implosion experiments are used to measure nTOF backgrounds in order to infer backscattered areal densities on OMEGA.

Deuterium↗
DOE OSTI2026

Interpreting AI for fusion: An application to plasma profile analysis for tearing mode stability

Artificial intelligence models have demonstrated strong predictive capabilities for various instabilities in fusion devices such as Tokamaks, including tearing modes (TM), edge localized modes, and disruptive events, but their opaque nature raises concerns about safety and trustworthiness when applied to fusion power plants. Here, we present a physics-based interpretation framework using a TM prediction model as a demonstration that is validated through a dedicated DIII-D TM avoidance experiment. By applying Shapley analysis, we identify how profiles such as rotation, temperature, and density contribute to the model's prediction of TM stability. Our analysis shows that in our experimental scenario, core electron temperature and rotation peaking play the primary role in TM stability, while density changes have smaller effects on stability. We show that off-axis ion temperature stabilizes TMs, suggesting that off-axis neutral beam heating can further stabilize this scenario. This work presents a generalizable ML-based event prediction methodology, from training to physics-driven interpretation, bridging the gap between physics understanding and opaque ML models.

Farre-Kaga, Hiro J. [Princeton Univ., NJ (United S↗
DOE OSTI2026

Combined measurements and interpretations of Higgs boson production and decay in proton–proton collisions at $\sqrt{s}$ = 13 TeV

Combined measurements of Higgs boson production and decay rates are reported, representing the most comprehensive study performed by the CMS Collaboration to date. The included analyses use proton–proton collision data recorded by the CMS experiment at $\sqrt{s}$ = 13 TeV from 2016 to 2018, corresponding to an integrated luminosity of 138 fb –1 . The statistical combination is based on analyses that measure the following decay channels: H → γγ, H → ZZ, H → WW, H → ππ, H → bb, H → μμ, and H → Zγ. Information in the events from each decay channel is used to target multiple Higgs boson production processes. Searches for invisible Higgs boson decays are also considered, as well as an analysis that measures off-shell Higgs boson production in the H → ZZ → 4$\ell$ ($\ell$ = e, μ) decay channel. The best fit inclusive signal yield is measured to be 1.014$^{+0.055}_{–0.053}$ times the standard model (SM) expectation, for a Higgs boson mass of 125.38 GeV. Measurements in kinematic regions defined by the simplified template cross section framework are also provided, as well as interpretations in the coupling modifier and SM effective field theory frameworks. The coupling modifier interpretation is further used to place constraints on various two-Higgs-doublet models. The results show good compatibility with the SM predictions for the majority of the measured parameters.

CMS↗
DOE OSTI2025

Massive νs through the CNN lens: interpreting the field-level neutrino mass information in weak lensing

Modern cosmological surveys probe the Universe deep into the nonlinear regime, where massive neutrinos suppress cosmic structure. Traditional cosmological analyses, which use the 2-point correlation function to extract information, are no longer optimal in the nonlinear regime, and there is thus much interest in extracting beyond-2-point information to improve constraints on neutrino mass. Quantifying and interpreting the beyond-2-point information is thus a pressing task. We study the field-level information in weak lensing convergence maps using convolution neural networks. We find that the network performance increases as higher source redshifts and smaller scales are considered — investigating up to a source redshift of 2.5 and ℓ max ≃ 10 4 — verifying that massive neutrinos leave a distinct effect on weak lensing. However, the performance of the network significantly drops after scaling out the 2-point information from the maps, implying that most of the field-level information can be found in the 2-point correlation function alone. We quantify these findings in terms of the likelihood ratio and also use Integrated Gradient saliency maps to interpret which parts of the map the network is learning the most from. We find that, in the absence of noise, the network extracts a similar amount of information from the most overdense and underdense regions. However, upon adding noise, the information in underdense regions is distorted as noise disproportionately washes out void-like structures.

Golshan, Malika [University of California, Berkele↗