MultiSolSegment: Trained Model Weights
The resulting model weights for MultiSolSegment to be loaded into the python code for other to use.
14 SOLAR ENERGY↗
SEARCH · Search NASA
Results for “model weights”
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.
Validation of an Erythema-Weighted UV Model Using Broadband Solar Irradiance Measurements From Eleven U.S. Sites: Preprint
Erythema-weighted UV solar irradiance (UV-E) has a potential impact on human health if the recommended maximum exposure times are exceeded. In spite of this, it is not measured at most sites that measure Global Horizontal Irradiance (GHI). However, since UV-E is highly correlated with GHI and total ozone content it can be estimated from this information with sufficient accuracy to assisst in public health recommendations. The Power Model (PM) provides a simple method to estimate the erythema-weighted UV irradiance (UV-E) from measured GHI, total ozone column and air mass. In this work, the performance of the PM method is assessed using high-quality data from 11 sites in the continental U.S. (part of SURFRAD and SOLRAD networks) and total ozone estimates publicly available from the MERRA-2 re-analysis database. A three year period (2021- 2023) at 1-minute frequency is considered. The results (for time aggregations of 5 and 60 minutes) show a high Pearson's correlations (> 0.99), consistently positive mean bias deviations (below 13%) and dispersions in the 8-19% range at all sites. Relative values are expressed in terms of the corresponding measurement mean. The performance indicators remain consistent across time resolutions (5 or 60 minutes), suggesting that the model's performance is robust and not significantly affected by short-term variability (which is captured by GHI). Spatial patterns reveal higher biases and RMSD in northern and eastern locations. These values represent a significant improvement over widely used satellite-based global UV-E estimates and open the possibility of using the PM with satellite-based GHI estimates for operational UV-E mapping over the contiguous U.S. territory.
14 SOLAR ENERGY↗
QCalEval: Benchmarking Vision-Language Models for Quantum Calibration Plot Understanding
Quantum computing calibration depends on interpreting experimental data, and calibration plots provide the most universal human-readable representation for this task, yet no systematic evaluation exists of how well vision-language models (VLMs) interpret them. We introduce QCalEval, the first VLM benchmark for quantum calibration plots: 243 samples across 87 scenario types from 22 experiment families, spanning superconducting qubits and neutral atoms, evaluated on six question types in both zero-shot and in-context learning settings. The best general-purpose zero-shot model reaches a mean score of 72.3, and many open-weight models degrade under multi-image in-context learning, whereas frontier closed models improve substantially. A supervised fine-tuning ablation at the 9-billion-parameter scale shows that SFT improves zero-shot performance but cannot close the multimodal in-context learning gap. As a reference case study, we release NVIDIA Ising Calibration 1, an open-weight model based on Qwen3.5-35B-A3B that reaches 74.7 zero-shot average score.
Cao, Shuxiang↗
Xanthos-Lake Dataset
The Xanthos-Lake v1.0 dataset provides the input data, trained machine-learning models, and simulation outputs needed to characterize lake water balance, snow and ice conditions, and mixing-layer temperature within the Xanthos global hydrological modeling framework. The dataset supports lake representation across a wide range of lake sizes and hydroclimatic conditions by combining xLSIM, a basin-specific machine-learning emulator of lake snow, ice, ice-cover fraction, and mixing-layer temperature, with the Xanthos-Lake water-balance model. The archive contains NetCDF datasets used to train and evaluate xLSIM, trained model weights, processed meteorological and lake-property inputs, and basin- and lake-category-specific simulation outputs. These materials are organized into four primary data groups, described below. Snowice_model_inputs: Contains the NetCDF input data used to train xLSIM. The xLSIM machine-learning framework uses three lake-based datasets. The meteorological forcing dataset provides monthly relative humidity, specific humidity, surface wind speed, maximum and minimum air temperature, downward longwave and shortwave radiation, snowfall, surface air pressure, and total precipitation. Lake surface area is included as an additional static predictor. The target-state dataset provides lake ice thickness, snow depth, snow cover, and lake mixing-layer temperature, while a companion lake-surface dataset provides the lake ice-cover fraction. Before training, ice thickness and snow depth are converted from meters to centimeters, mixing-layer temperature is converted from kelvin to degrees Celsius and constrained to nonnegative values, and ice-cover fraction is converted from a fraction to a percentage. The predictor variables are normalized using statistics calculated across the selected lakes and time steps. Snowice_model_outputs: Contains the NetCDF outputs generated by xLSIM. For each basin, xLSIM produces a file containing observed and predicted lake-state variables for the training, validation, and testing periods. The modeled variables include lake ice thickness, snow depth, snow cover, mixing-layer temperature, and lake ice-cover fraction. For basins without a sufficiently persistent snow-and-ice signal, the emulator predicts only mixing-layer temperature. The outputs also include training and validation loss histories, the selected model configuration, identifiers of the lakes used in training, and SHAP-based feature-importance information at the global, lake, and seasonal-regime levels. The trained machine-learning model weights are provided separately within the dataset archive. Together, these files support model evaluation and subsequent coupling with the Xanthos-Lake water-balance framework. XanthosLAKES: Contains the NetCDF input data used by the Xanthos-Lake framework. Monthly meteorological inputs include relative and specific humidity, downward shortwave and longwave radiation, mean, maximum, and minimum air temperature, wind speed, precipitation, snowfall, and surface air pressure. Static lake-property datasets provide lake identifiers, geographic locations, surface area, volume, mean depth, elevation, drainage area, fetch, outlet-routing information, and associated Xanthos grid-cell attributes. Separate bathymetric datasets provide the coefficients of the area–depth and volume–depth relationships for each aggregated lake unit. GLEV-based records provide observed lake surface area and evaporation data used to initialize lake states, define reference conditions, and calibrate and evaluate the model. Xanthos-Lake Outputs: Contains the basin- and lake-category-specific NetCDF outputs generated by Xanthos-Lake. Monthly variables include lake surface area, storage volume, outlet discharge, evaporation rate, evaporation volume, lake–groundwater exchange, lake inflow, ice thickness, snow depth, snow-cover fraction, ice-cover fraction, and mixing-layer temperature. The files also contain lake-specific calibration and validation statistics, including normalized root-mean-square error, mean absolute error, Nash–Sutcliffe efficiency, Kling–Gupta efficiency, and percent bias. Stored calibrated and derived parameters include the weir discharge coefficient, fractional freeboard, groundwater exchange coefficient, reference water level, corresponding reference surface area and storage volume, weir-width adjustment factor, and the fraction of routed inflow entering the lake. Basin identifiers, lake category, simulation period, calibration and validation periods, and parameter-schema information are retained as NetCDF metadata.
Abeshu, Guta [Pacific Northwest National Laborator↗
Spectra-to-exposure conversion using polynomial response models for gamma-ray field characterization
Accurate measurement of exposure rate from gamma-ray spectral data remains a critical challenge during radiological emergency response operations. Conventional methods rely on pre-defined static conversion factors derived from fixed geometries and isotopic compositions, which often fail to capture real-world environmental variability. This study presents a generalized approach as a "next-step" for converting gamma-ray spectral data into exposure rate using polynomial response models. The method introduces a flexible weighting scheme based on the in-situ detector response to distributed sources, enabling a pathway towards improved correspondence between measured spectra and "ground-truth" exposure rates. Experimental data from sodium iodide NaI(Tl) detectors were used to validate the approach as, at least equivalent to the current count-to-exposure method employed in emergency response CONOPS. Results show that the polynomial weighting model is sufficiently equal to the count-to-exposure method and may help improve accuracy given its adaptability to real-world conditions.
61 RADIATION PROTECTION AND DOSIMETRY↗
pvcracks: trained VAE model
The resulting model weights for the variational autoencoder for solar cell crack parametrization to be loaded into the python code for other to use
14 SOLAR ENERGY↗
The Double-edged Sword of Data-driven Super-Resolution: Adversarial Super-resolution Models
Data-driven super-resolution (SR) methods are often integrated into imaging pipelines as preprocessing steps to improve downstream tasks such as classification and detection. However, these SR models introduce a previously unexplored attack surface into imaging pipelines. In this paper, we present AdvSR, a framework demonstrating that adversarial behavior can be embedded directly into SR model weights during training, requiring no access to inputs at inference time. Unlike prior attacks that perturb inputs or rely on backdoor triggers, AdvSR operates entirely at the model level. By jointly optimizing for reconstruction quality and targeted adversarial outcomes, AdvSR produces models that appear benign under standard image quality metrics while inducing downstream misclassification. We evaluate AdvSR on three SR architectures (SRCNN, EDSR, SwinIR) paired with a YOLOv11 classifier and demonstrate that AdvSR models can achieve high attack success rates with minimal quality degradation. These findings highlight a new model-level threat for imaging pipelines, with implications for how practitioners source and validate models in safety-critical applications.
Sullivan, Haley [ORNL] (ORCID:0000000274069217)↗
SymbolNet: neural symbolic regression with adaptive dynamic pruning for compression
Abstract Compact symbolic expressions have been shown to be more efficient than neural network (NN) models in terms of resource consumption and inference speed when implemented on custom hardware such as field-programmable gate arrays (FPGAs), while maintaining comparable accuracy (Tsoi et al 2024 EPJ Web Conf. 295 09036). These capabilities are highly valuable in environments with stringent computational resource constraints, such as high-energy physics experiments at the CERN Large Hadron Collider. However, finding compact expressions for high-dimensional datasets remains challenging due to the inherent limitations of genetic programming (GP), the search algorithm of most symbolic regression (SR) methods. Contrary to GP, the NN approach to SR offers scalability to high-dimensional inputs and leverages gradient methods for faster equation searching. Common ways of constraining expression complexity often involve multistage pruning with fine-tuning, which can result in significant performance loss. In this work, we propose S y m b o l N e t , a NN approach to SR specifically designed as a model compression technique, aimed at enabling low-latency inference for high-dimensional inputs on custom hardware such as FPGAs. This framework allows dynamic pruning of model weights, input features, and mathematical operators in a single training process, where both training loss and expression complexity are optimized simultaneously. We introduce a sparsity regularization term for each pruning type, which can adaptively adjust its strength, leading to convergence at a target sparsity ratio. Unlike most existing SR methods that struggle with datasets containing more than O ( 10 ) inputs, we demonstrate the effectiveness of our model on the LHC jet tagging task (16 inputs), MNIST (784 inputs), and SVHN (3072 inputs).
Tsoi, Ho Fung (ORCID:0000000225502184)↗
A comparative study of multimodal data fusion strategies for planetary spectroscopy
Integrating heterogeneous data sources can improve scientific inference when different modalities capture complementary information, but doing so is challenging in high-dimensional, small-sample settings. In spectroscopy for planetary exploration, Laser-Induced Breakdown Spectroscopy (LIBS), Raman Spectroscopy (Raman), Visible Infrared Spectroscopy (VISIR), and Mid-Infrared Spectroscopy (MIR) each examine different aspects of composition and mineralogy, raising fundamental questions about when and how data fusion improves predictive performance. Using a Mars-relevant set of geologic standards with measurements from all four modalities, we present a rigorous systematic evaluation of four data fusion strategies: low-level (data) fusion, mid-level (feature) fusion, high-level (decision) fusion, and residual-boosting (sequential) fusion. We assess performance in predicting oxide composition via nested cross-validation and corrected significance testing to evaluate whether data fusion improves upon single-modality baselines. We show that data fusion does not uniformly improve accuracy, and that observed gains are modest, oxide-dependent, and sensitive to modality and model structure. To move beyond aggregate accuracy metrics, we use model coefficients, permutation importance, and residual gain analysis to examine how the fusion models weight individual modalities and to identify patterns of apparent complementarity or redundancy. Though focused on spectroscopy for planetary exploration, our framework for data fusion evaluation and interpretation extends to other scientific domains with heterogeneous and scarce data and provides a principled approach evaluating data fusion strategies, interpreting modality contributions, and understanding tradeoffs among data fusion strategies.
97 MATHEMATICS AND COMPUTING↗
FLYING SERVING: On-the-Fly Parallelism Switching for Large Language Model Serving
Production LLM serving must simultaneously deliver high throughput, low latency, and sufficient context capacity under non-stationary traffic and mixed request requirements. Data parallelism (DP) maximizes throughput by running independent replicas, while tensor parallelism (TP) reduces per-request latency and pools memory for long-context inference. However, existing serving stacks typically commit to a static parallelism configuration at deployment; adapting to bursts, priorities, or long-context requests is often disruptive and slow. We present Flying Serving, a vLLM-based system that enables online DP-TP switching without restarting engine workers. Flying Serving makes reconfiguration practical by virtualizing the state that would otherwise force data movement: (i) a zero-copy Model Weights Manager that exposes TP shard views on demand, (ii) a KV Cache Adaptor that preserves request KV state across DP/TP layouts, (iii) an eagerly initialized Communicator Pool to amortize collective setup, and (iv) a deadlock-free scheduler that coordinates safe transitions under execution skew. Across three popular LLMs and realistic serving scenarios, Flying Serving improves performance by up to 4.79 × under high load and 3.47 × under low load while supporting latency- and memory-driven requests.
Gao, Shouwei [ORNL]↗
Practical Probabilistic Programming
Recent advances in probabilistic programming languages (PPLs) have provided the capability for exact inference: computing a closed-form probability distribution for a given probabilistic program. In particular, the new language Roulette uses a language oriented programming (LOP) approach, wherein analysts build new programming languages on top of a set of primitives provided by Roulette, which then translates these structures into a weighted model counting problem which can be solved by automated reasoning tools. However, because Roulette provides few convenience features, developing these new languages is challenging even for expert users. We developed a standard library of common probability functions for Roulette with the goal of improved usability. This included approximation of continuous probability density functions using discrete probability mass functions. We demonstrated this approach by modeling a cosmic ray striking a RAM controller. We found that Roulette provides a powerful interface for highly expressive probabilistic programs to be generated. In collaboration with the NNSA Advanced Simulation and Computing program, which resulted in development of a tool called Circulette, we were able to model complex circuits expressed in Verilog using probabilistic programs with an expressivity not previously possible. Our research question that motivated the development of a Roulette standard library was to determine whether non-experts could use a PPL to model relevant problems regarding radiation effects on microelectronics. This standard library improved the expressivity of Roulette by implementing common probability density functions, mathematical operators on distributions, and support for empirical distributions. While Roulette is a powerful modeling language, the untyped, LOP approach makes error messages difficult to understand and requires expert aid. We recommend further research on Roulette, especially with its error messages, to enable improved usability. At the same time, this project demonstrated that for users familiar with Roulette and the LOP approach, Roulette provides powerful new capabilities that can be integrated with other Sandia modeling capabilities.
97 MATHEMATICS AND COMPUTING↗
Archi: Agentic Operations at the CMS Experiment
We present Archi, an open-source, end-to-end framework for scientific collaborations that combines the systematic ingestion and organization of heterogeneous data sources with the deployment of configurable, private, and extensible agents that retrieve and reason over them. An instance of Archi has been deployed for the Computing Operations team of the CMS experiment at CERN's LHC since February 2026 as a support agent for technical operators, offering retrieval and analysis capabilities by combining documentation, historical data, and live monitoring systems. We evaluate the system on operator feedback and a question set collected from production usage, graded by human and automated panels. The system proves effective at operational tasks, resolving real-world queries posed by CMS operators. We also observe that locally-hosted, open-weight models perform competitively, enabling fully private management of sensitive data.
Lugato, Pietro [MIT; CERN]↗
Scaling open-weight large language models for hydropower regulatory information extraction: A systematic analysis
Information extraction from regulatory and technical documents using large language models (LLMs) involves practical trade-offs between extraction quality and computational cost. We evaluate eight open-weight LLMs spanning 0.6B–70B parameters on hydropower licensing documents and report deployment-oriented evidence under a unified extraction schema and evaluation protocol. Across the model set, we observe clear scale-dependent trends in both baseline extraction quality and the effectiveness of reflective reasoning (self-checking) under our fixed-prompt, no-augmentation setting. Mid-scale models often provide a favorable balance of accuracy and efficiency, whereas the smallest models show limited or inconsistent gains from the reasoning variants tested. Larger models achieve the highest overall F1 scores but incur substantially greater compute and infrastructure requirements. We further find that reliability failure modes can distort conventional metrics in this domain: in particular, high recall can coincide with systematic extraction errors when models fabricate values for fields that are absent from the source text, underscoring the importance of conservative null handling and evidence-grounded evaluation. Overall, our study provides a reproducible resource–performance comparison for open-weight LLM-based extraction in hydropower regulatory documentation and offers practical guidance for model selection under different deployment constraints.
Evaluation protocol↗
Model-free Rayleigh weight from x-ray Thomson scattering measurements
X-ray Thomson scattering (XRTS) has emerged as a powerful tool for the diagnostics of matter under extreme conditions. In principle, it gives one access to important system parameters such as the temperature, density, and ionization state, but the interpretation of the measured XRTS intensity usually relies on theoretical models and approximations. In this context, a key property is given by the Rayleigh weight that describes the electronic localization around the ions. Here, we show that it is possible to extract the Rayleigh weight directly from the experimental data without the need for any model calculations or simulations. As a practical application, we consider an experimental measurement of strongly compressed Be at the National Ignition Facility [Döppner et al., Nature 618, 270–275 (2023)]. We demonstrate that experimental results for the Rayleigh weight open up new avenues for the interpretation of XRTS experiments by matching the measurement with ab initio simulations such as density functional theory or path integral Monte Carlo. Interestingly, this new procedure leads to significantly lower density compared to previously used chemical models.
70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
Context Modulation Enables Model Robustness to Weight Noise in Recurrent Spiking Neural Networks
Explore the source record for details and available documents.
Luca, Sarah Katherine [Sandia National Laboratorie↗
Ab Initio Polariton Spectra of ZnTPP Molecules Collectively Coupled Inside an Optical Cavity
Exciton-polaritons are quasi-particles formed by the quantum mechanical hybridization of electronic and photonic excitations. Despite extensive investigations, a fundamental understanding of molecular polariton spectra and the polariton delocalization from an ab initio theoretical perspective remains elusive. We simulate experimentally measured linear transmission spectroscopy of many Zinc(II) tetraphenylporphyrin (ZnTPP) molecules collectively coupled to a cavity from first principles. Our theoretical approach incorporates many low-lying electronic excitations in ZnTPP molecules, as well as collective light-matter couplings between ZnTPP and the quantized radiation modes, both of which are shown to be the key to accurately recovering the experimental spectra. We further analyzed to what extent the polariton and dark states are delocalized over many molecules, for the first time, using fully ab initio descriptions of the molecules. We finally investigate the line width as a function of detuning, providing new theoretical insights into the experimentally observed motional narrowing behavior. Our work presents first-ofits- kind theoretical studies on molecular polariton spectra, offering a new perspective on molecular polariton formation in realistic ab initio molecular systems whose rich, many-state nature provides spectral features enabled by the high density of electronic states beyond simple quantum optics models.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Optimal Methods for Estimating Cactus Pear Biomass Using Cladode Dimensions of Morphologically Diverse Accessions
Current allometric methods for photosynthetic-stem (cladode) plants, such as cactus pear (Opuntia spp.), require refinement to be used in field settings in which diverse accessions are grown. We analysed cladode dimensional data using 14 accessions representing four species and two hybrids to quantify statistically significant morphological differences among accessions and derived cross-accession models to approximate cladode fresh weight. A Box model using cladode dimensions (e.g., length, width, thickness and diameter) and factorial combinations of these measures (e.g., length*width*thickness*diameter vs. fresh weight) resulted in the highest coefficient of determination (R 2 = 0.95 general fit) across all accessions for estimating fresh weight along with parsimony estimates using the Schwarz–Bayes Criterion (SBC), which assesses the most consistent performance on individual accessions. A Fitting-box modelling approach used the measured cladode area captured using ImageJ (R 2 = 0.93 general fit). Lastly, an Elliptical model used an elliptical approximation for the measured area and performed well over all accessions (R 2 = 0.94 general fit) while avoiding extensive manual measurements. These models meet or exceed the performance of previously published approaches when applied across morphologically diverse accessions, providing efficient tools for nondestructive estimation of cactus pear biomass under the conditions tested.
Opuntia↗
Geographical Insights into Suicide Mortality Through Spatial Machine Learning
Suicide mortality is a leading cause of death in the United States, with an upward trend that emphasizes its significance as a public health issue. Previous research has employed global models like ordinary least squares (OLS) regression and local models such as geographically weighted regression (GWR). While local models are useful for analyzing spatial variations in suicide mortality, they share limitations with traditional global models, particularly about their inability to handle multi-collinearity and non-linear relationships. Machine learning approaches, like random forests (RF), can address some of these limitations but often fail to account for spatial variability. This gap highlights the need for spatial ML models specifically designed to tackle suicide mortality. This research seeks to fill this void by using a geographically weighted random forest model (GWRF) to examine the associations between county-level suicide mortality in the U.S. from 2010 to 2020 and various social and environmental determinants of health. A key aspect of our methodology is disciplined feature selection, which reduces the pool of explanatory variables by about 90%. This refinement enhances the explanatory power of both global (R2 improved from 0.59 to 0.67) and local (R2 improved from 0.64 to 0.67) RF models while reducing their run times. An analysis of the importance scores for these selected features reveals that the drivers of suicide mortality vary by context. Thus, to effectively address regional disparities and inform targeted public health interventions, a holistic approach that incorporates multiple county-level characteristics is essential.
Lebakula, Viswadeep [ORNL] (ORCID:0000000152935914↗