Search NASA⌕ Search

SEARCH · Search NASA

Results for “disentangled representations”

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.

16 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Uncovering obscured phonon dynamics from powder inelastic neutron scattering using machine learning

The study of phonon dynamics is pivotal for understanding material properties, yet it faces challenges due to the irreversible information loss inherent in powder inelastic neutron scattering spectra and the limitations of traditional analysis methods. In this study, we present a machine learning framework designed to reveal obscured phonon dynamics from powder spectra. Using a variational autoencoder, we obtain a disentangled latent representation of spectra and successfully extract force constants for reconstructing phonon dispersions. Notably, our model demonstrates effective applicability to experimental data even when trained exclusively on physics-based simulations. The fine-tuning with experimental spectra further mitigates issues arising from domain shift. Analysis of latent space underscores the model’s versatility and generalizability, affirming its suitability for complex system applications. Furthermore, our framework’s two-stage design is promising for developing a universal pre-trained feature extractor. This approach has the potential to revolutionize neutron measurements of phonon dynamics, offering researchers a potent tool to decipher intricate spectra and gain valuable insights into the intrinsic physics of materials.

domain adaptation↗
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 OSTI2024

Soil Moisture‐Cloud‐Precipitation Feedback in the Lower Atmosphere From Functional Decomposition of Satellite Observations

Abstract The feedback of topsoil moisture (SM) content on convective clouds and precipitation is not well understood and represented in the current generation of weather and climate models. Here, we use functional decomposition of satellite‐derived SM and cloud vertical profiles (CVP) to quantify the relationship between SM and the vertical distribution of cloud water in the central US. High‐dimensional model representation is used to disentangle the contributions of SM and other land‐surface and atmospheric variables to the CVP. Results show that the sign and strength of the SM‐cloud‐precipitation feedback varies with cloud height and time lag and displays a large spatial variability. Positive anomalies in antecedent 7‐hr SM and land‐surface temperature enhance cloud reflectivity up to 4 dBZ in the lower atmosphere about 1–3 km above the surface. Our approach presents new insights into the SM‐cloud‐precipitation feedback and aids in the diagnosis of land‐atmosphere interactions simulated by weather and climate models.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2026

CAD-VAE: Leveraging Correlation-Aware Latents for Comprehensive Fair Disentanglement

While deep generative models have significantly advanced representation learning, they may inherit or amplify biases and fairness issues by encoding sensitive attributes alongside predictive features. Enforcing strict independence in disentanglement is often unrealistic when target and sensitive factors are naturally correlated. To address this challenge, we propose CAD-VAE(Correlation-Aware Disentangled VAE), which introduces a correlated latent code to capture the information shared between the target and sensitive attributes. Given this correlated latent, our method effectively separates over-lapping factors without extra domain knowledge by directly minimizing the conditional mutual information between target and sensitive codes. A relevance-driven optimization strategy refines the correlated code by efficiently capturing essential correlated features and eliminating redundancy. Extensive experiments on benchmark datasets demonstrate that CAD-VAE produces fairer representations, realistic counterfactuals, and improved fairness-aware image editing.

Ma, Chenrui [University of California Irvine]↗
DOE OSTI2024

Physics and chemistry from parsimonious representations: image analysis via invariant variational autoencoders

Electron, optical, and scanning probe microscopy methods are generating ever increasing volume of image data containing information on atomic and mesoscale structures and functionalities. This necessitates the development of the machine learning methods for discovery of physical and chemical phenomena from the data, such as manifestations of symmetry breaking phenomena in electron and scanning tunneling microscopy images, or variability of the nanoparticles. Variational autoencoders (VAEs) are emerging as a powerful paradigm for the unsupervised data analysis, allowing to disentangle the factors of variability and discover optimal parsimonious representation. Here, we summarize recent developments in VAEs, covering the basic principles and intuition behind the VAEs. The invariant VAEs are introduced as an approach to accommodate scale and translation invariances present in imaging data and separate known factors of variations from the ones to be discovered. We further describe the opportunities enabled by the control over VAE architecture, including conditional, semi-supervised, and joint VAEs. Several case studies of VAE applications for toy models and experimental datasets in Scanning Transmission Electron Microscopy are discussed, emphasizing the deep connection between VAE and basic physical principles. Python codes and datasets discussed in this article are available at https://github.com/saimani5/VAE-tutorials and can be used by researchers as an application guide when applying these to their own datasets.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Enhancing Interpretability in Generative Modeling: Statistically Disentangled Latent Spaces Guided by Generative Factors in Scientific Datasets

This study addresses the challenge of statistically extracting generative factors from complex, high-dimensional datasets in unsupervised or semi-supervised settings. We investigate encoder-decoder-based generative models for nonlinear dimensionality reduction, focusing on disentangling low-dimensional latent variables corresponding to independent physical factors. Introducing Aux-VAE, a novel architecture within the classical Variational Autoencoder framework, we achieve disentanglement with minimal modifications to the standard VAE loss function by leveraging prior statistical knowledge through auxiliary variables. These variables guide the shaping of the latent space by aligning latent factors with learned auxiliary variables. We validate the efficacy of Aux-VAE through comparative assessments on multiple datasets, including astronomical simulations.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

NuGraph2 with context-aware inputs: physics-inspired improvements in semantic segmentation

Graph neural networks have recently shown strong promise for event reconstruction tasks in Liquid Argon Time Projection Chambers, yet their performance remains limited for underrepresented classes of particles, such as Michel electrons. In this work, we investigate physics-informed strategies to improve semantic segmentation within the NuGraph2 architecture. We explore three complementary approaches: (i) enriching the input representation with context-aware features derived from detector geometry and track continuity, (ii) introducing auxiliary decoders to capture class-level correlations, and (iii) incorporating energy-based regularization terms motivated by Michel electron energy distributions. Experiments on MicroBooNE public datasets show that physics-inspired feature augmentation yields the largest gains, particularly boosting Michel electron precision and recall by disentangling overlapping latent space regions. In contrast, auxiliary decoders and energy-regularization terms provided limited improvements, partly due to the hit-level nature of NuGraph2, which lacks explicit particle- or event-level representations. Our findings highlight that embedding physics context directly into node-level inputs is more effective than imposing task-specific auxiliary losses, and suggest that future hierarchical architectures such as NuGraph3, with explicit particle- and event-level reasoning, will provide a more natural setting for advanced decoders and physics-based regularization. The code for this work is publicly available on Github at https://github.com/vitorgrizzi/nugraph_phys/tree/main_phys.

Other Experiments↗
DOE OSTI2026

Attention-based explainability for structure–property relationships

Machine learning methods are emerging as a universal paradigm for constructing correlative structure–property relationships in materials science based on multimodal characterization. However, this necessitates the development of methods for the physical interpretability of the resulting correlative models. Here, we demonstrate the potential of attention-based neural networks for revealing structure–property relationships and the underlying physical mechanisms, using the ferroelectric properties of PbTiO3 thin films as a case study. Through the analysis of attention scores, we disentangle the influence of distinct domain patterns on the polarization switching process. The attention-based Transformer model is explored both as a direct interpretability tool and as a surrogate for explaining representations learned via unsupervised machine learning, enabling the identification of physically grounded correlations. We compare attention-derived interpretability scores with classical SHapley Additive exPlanations analysis and show that, in contrast to applications in natural language processing, attention mechanisms in materials science exhibit high efficiency in highlighting meaningful structural features.

Slautin, Boris [Independent Researcher]↗
DOE OSTI2025

The Two Arctic Wintertime Boundary Layer States: Disentangling the Role of Cloud and Wind Regimes in Reanalysis and Observations During MOSAiC

The wintertime central Arctic atmosphere comprises a radiatively clear and a radiatively opaque state, which are linked to synoptic forcing and mixed-phase clouds. Weather and climate models often lack process representations surrounding these states, but prior work mostly treated the problem as an aggregate of synoptic conditions, resulting in partially overlapping biases. Here, we disaggregate the Arctic states and confront ERA5 reanalysis with observations from the MOSAiC campaign over the central Arctic sea ice during winter 2019/2020. Low-level winds and liquid water path (LWP) are combined to derive different synoptic classes. Results show that the clear state is primarily formed by weak/moderate winds and the absence of liquid-bearing clouds, while strong winds and enhanced LWP primarily form the radiatively opaque state. ERA5 struggles to reproduce these basic statistics, shows too weak sensitivity of thermal radiation to synoptic forcing, and overestimates thermal radiation for similar LWP amounts. The latter is caused by a warm bias, which has a pronounced inversion structure and is largest in clear and calm conditions. Under strong synoptic forcing, the warm bias is constant with height and discrepancies in mixed-phase cloud altitude appear. Separating synoptic conditions is regarded as useful for process-oriented evaluation of the Arctic troposphere in models.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2026

Locally purified maximally mixed states at scale: Entanglement pruning and symmetries

Locally Purified Density Operators (LPDOs) are state-of-the-art tensor network ansatze candidates that efficiently represent mixed quantum states at scale. However, given their non-uniqueness, their representational complexity is generally sub-optimal in practical computations. Here, in this work we perform a comprehensive numerical and analytical analysis and resolve this issue in the experimentally relevant limit where noise depolarizes the density operator into a maximally mixed state. To resolve the sub-optimality issue, we analyze two numerical tools, one analytic method, and detail the relations between them. The numerical tools used are fidelity-preserving truncations and isometric gauge transformations leveraging Riemannian optimizations over entropic objective functions. In addition, by invoking the injectivity and symmetry constraints of the maximally mixed LPDO, we also present analytical closed-form expressions for the disentangler and discuss their relation to numerical optimizers. Further, away from the maximally mixed state, our simulations highlight how the truncation threshold smoothly interpolate, as a function of depolarization, between established matrix product results and our new results. Our work shows how, by minimizing the resources required to represent key states of practical interest in experiment, the efficiency of tensor network algorithms can be substantially increased. This paves the path for uncovering tensor network’s fundamental scalability limits and latent potential in representing the wide locus of mixed quantum states that are accessible on near-term quantum devices.

Gangapuram, Amit Jamadagni [Oak Ridge National Lab↗
DOE OSTI2024

Physical discovery in representation learning via conditioning on prior knowledge

Recent advances in electron, scanning probe, optical, and chemical imaging and spectroscopy yield bespoke data sets containing the information of structure and functionality of complex systems. In many cases, the resulting data sets are underpinned by low-dimensional simple representations encoding the factors of variability within the data. The representation learning methods seek to discover these factors of variability, ideally further connecting them with relevant physical mechanisms. However, generally, the task of identifying the latent variables corresponding to actual physical mechanisms is extremely complex. Here, we present an empirical study of an approach based on conditioning the data on the known (continuous) physical parameters and systematically compare it with the previously introduced approach based on the invariant variational autoencoders. The conditional variational autoencoder (cVAE) approach does not rely on the existence of the invariant transforms and hence allows for much greater flexibility and applicability. Interestingly, cVAE allows for limited extrapolation outside of the original domain of the conditional variable. However, this extrapolation is limited compared to the cases when true physical mechanisms are known, and the physical factor of variability can be disentangled in full. We further show that introducing the known conditioning results in the simplification of the latent distribution if the conditioning vector is correlated with the factor of variability in the data, thus allowing us to separate relevant physical factors. We initially demonstrate this approach using 1D and 2D examples on a synthetic data set and then extend it to the analysis of experimental data on ferroelectric domain dynamics visualized via piezoresponse force microscopy.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2025

A Variational Autoencoder Model Toward Molecular Structure Representation Learning of Fuels

Here, in this work, a Variational Autoencoder (VAE)-based data-driven modeling framework is developed with the overarching goal of enabling fuel design. The VAE model is trained on a large dataset with several chemical species to learn a compressed latent space molecular representation. Chemical structure in the form of Simplified Molecular Input Line Entry System (SMILES) string is fed as input, encoded into the VAE latent space, and decoded back to the SMILES string using Long Short-Term Memory (LSTM) networks. Complexities of the VAE training loss function are thoroughly examined by varying the weightage (beta (𝜷) parameter) of the latent space regularization term, thereby assessing the balance between reconstruction accuracy and validity, and focusing on both accurate molecular structure reconstruction and latent space consistency. Two different strategies for 𝜷 variation are evaluated: linear annealing and cyclic annealing. In addition, the impact of total correlation adjustment and hierarchical priors is also studied with regard to the balance between reconstruction fidelity and latent space regularization, and potential issues such as posterior collapse, over-regularization, and poor disentanglement of latent variables. Overall, the best performance of the model is achieved with hierarchical priors and incrementally increasing 𝜷 from 0 to a threshold value of 0.25 over 75 epochs. The generative VAE model can be readily coupled with Quantitative Structure–Property Relationship (QSPR) analysis to develop an integrated end-to-end framework for fuel-property prediction and molecular design of novel promising fuels.

fuel design↗
DOE OSTI2026

Characterization of carbonaceous aerosols during TRACER-CAT

Absorbing aerosols (AA) have an important impact on the global radiation budget and cloud properties. The composition and properties of AA can vary substantially throughout the atmosphere, depending on the particle source and the influence of chemical aging. Uncertainties associated with the radiative effects of AA remain substantial. A key contributor to this uncertainty is understanding the extent to which coatings in general, and water uptake especially, alters absorption by AA particles and how this depends on particle composition. We deployed new and existing experimental tools during the Tracking Aerosol Convection Interactions Experiment (TRACER) campaign in Houston, TX as part of the Carbonaceous Aerosols Thrust (CAT) to provide detailed characterization of aerosol optical, chemical, and physical properties. Our TRACER-CAT measurements complemented and expanded on the planned TRACER instrumentation, allowing for more detailed characterization of aerosol properties of relevance to cloud development (a core focus of TRACER), such as the composition of particles that can act as cloud condensation nuclei, than would otherwise be available. Our measurements have allowed for assessment of the relationship(s) between AA optical properties (with a focus on absorption) and the chemical and physical characteristics (including the mixing state of black carbon (BC) containing particles). These field observations occurred in collaboration with Los Alamos National Laboratory in summer 2022 during the TRACER intensive operating period. The instrumentation we co-deployed provided for measurement of (i) multi-wavelength dry aerosol absorption, scattering, and extinction, (ii) the size-dependent composition and abundance of sub-micron aerosol, differentiating between those particles that do and do not contain BC, (iii) BC-specific concentrations and size distributions, (iv) particle size, and (v) the first field measurements at an ARM site of the influence of RH on multi-wavelength absorption by ambient AA. We have leveraged the natural variability of the atmosphere and of aerosol sources in the Houston region to (i) specifically disentangle contributions to light absorption from BC, absorbing organic carbon (brown carbon), and coatings on BC, (ii) characterize the mixing state of BC and assess the factors that give rise to compositional differences between BC-containing and BC-free aerosol, and (iii) establish how water uptake influences absorption and how any such effect depends on particle composition and BC mixing state. Overall, our study contributed to the mission of the Atmospheric System Research program in multiple ways. Through the deployment of complementary, advanced instrumentation for characterization of a wide range of aerosol properties our work helped to maximize the scientific impact of the TRACER campaign. Our work also allowed for development of new insights into the relationship(s) between aerosol composition, hygroscopicity, and the mixing state of BC with aerosol optical properties. Through this, our work has provided knowledge that can improve understanding and model representation of aerosol processes as they affect the Earth’s radiation budget.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

Towards a liana plant functional type for vegetation models

Lianas (woody climbers) are crucial components of tropical forests and they have been increasingly recognized to have profound effects on tropical forest carbon dynamics. Despite their importance, lianas' representation in vegetation models remains limited, partly due to the complexity of liana-tree dynamics and the diversity in liana life history strategies. This paper provides a comprehensive review of advances and challenges for mechanistically representing lianas in forest ecosystem models and a proposed path towards effectively representing lianas in these models. Defining a liana plant functional type is a significant challenge because of the high morphological and physiological diversity amongst liana species, and because of their structural association with trees. Here, we identify critical liana traits that likely should contribute to establishing a liana plant functional type, along with key processes to properly represent lianas in ecosystem models. Subsequently, we discuss a variety of possible liana implementation strategies with their associated strengths, limitations, computational costs and data requirements. A fundamental redesign of the tree-centric demographic vegetation models seems appropriate to accommodate the unique growth and competition strategies of lianas. We illustrate the potential of such models with a single-site case study where we disentangle putative mechanisms of liana increasing abundance. Furthermore, we underscore the critical need for comprehensive liana demographic and functional data (including long-term, physiological, and pantropical observations) for the qualitative implementation and evaluation in the proposed modeling efforts. Currently, there is a scarcity of liana data and the data that do exist have a neotropical bias. We finally introduce a new liana functional trait database that can centralize existing liana trait data, incentivize improved data gathering and thus facilitate model development and scientific analyses.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Disentangling the Impacts of Microtopography and Shrub Distribution on Snow Depth in a Subarctic Watershed: Toward a Predictive Understanding of Snow Spatial Variability

Snow plays a critical role in carbon cycling, vegetation dynamics, and permafrost hydrology at high latitudes by influencing surface energy exchange. Predicting snow distribution patterns is essential for understanding the evolution of Arctic ecosystems, yet scaling process-level knowledge to landscape predictions remains challenging. Here, we analyze snow depth (2019 and 2022), terrain elevation, and vegetation height from a watershed on the Seward Peninsula, Alaska, to examine how topography and shrubs shape snow redistribution across spatial scales. We find that snow depth is strongly coupled to terrain at scales below ∼60 m but becomes increasingly decoupled at larger scales. The topographic model of snow depth variation, which transforms terrain data to align with these scale-dependent snow patterns, is well correlated with local snow depth variations (linear fit R 2 > 0.5 for 85% of 100-m patches). A machine learning reconstruction of shrub canopy snow trapping reveals a simple exponential relationship between canopy structure and snow accumulation ( R 2 = 0.59), highlighting the combined influence of topography and vegetation on snow distribution. Together, these empirical relationships capture much of the observed snow variability in the watershed ( R 2 = 0.49, root mean square error (RMSE) = 30 cm), though systematic limitations persist in areas of strong scour and at coarser scales where wind-terrain interactions are more complex. These findings provide a framework for more efficient snow depth prediction and offer insights to improve snow-vegetation feedback representation in Earth System Models.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Disentangling the Impacts of Microtopography and Shrub Distribution on Snow Depth in a Subarctic Watershed: Toward a Predictive Understanding of Snow Spatial Variability: Supporting Data and Code

This repository contains R code and associated datasets for reproducing the analysis described in the manuscript titled “Disentangling the Impacts of Microtopography and Shrub Distribution on Snow Depth in a Subarctic Watershed: Toward a Predictive Understanding of Snow Spatial Variability” (DOI: 10.1029/2024JG008604). The provided scripts facilitate a comprehensive analysis of snow depth variability influenced by microtopography and vegetation distribution in a subarctic watershed. Included datasets are high-resolution spatial maps of snow depth, terrain elevation, vegetation height, and distance from shrubs taller than 1 meter, all formatted as text files (.txt). These data are fully describe in doi:10.15485/2316038. Users can adapt the provided R scripts to accommodate different data formats or larger spatial domains, noting that some output files may require modification due to their size.The code includes implementations for boosted regression tree analysis adapted from methods outlined in Elith et al. (2008). Users interested in understanding or modeling landscape-scale snow distribution patterns, particularly in Arctic or subarctic ecosystems, will find this package useful. The Next-Generation Ecosystem Experiments: Arctic (NGEE Arctic), was a research effort to reduce uncertainty in Earth System Models by developing a predictive understanding of carbon-rich Arctic ecosystems and feedbacks to climate. NGEE Arctic was supported by the Department of Energy's Office of Biological and Environmental Research.The NGEE Arctic project had two field research sites: 1) located within the Arctic polygonal tundra coastal region on the Barrow Environmental Observatory (BEO) and the North Slope near Utqiagvik (Barrow), Alaska and 2) multiple areas on the discontinuous permafrost region of the Seward Peninsula north of Nome, Alaska.Through observations, experiments, and synthesis with existing datasets, NGEE Arctic provided an enhanced knowledge base for multi-scale modeling and contributed to improved process representation at global pan-Arctic scales within the Department of Energy's Earth system Model (the Energy Exascale Earth System Model, or E3SM), and specifically within the E3SM Land Model component (ELM).

54 ENVIRONMENTAL SCIENCES↗