Search NASA⌕ Search

SEARCH · Search NASA

Results for “Reasoning AI”

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.

86 records · Page 5
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Augmenting machine learning of Grad–Shafranov equilibrium reconstruction with Green's functions

This work presents a method for predicting plasma equilibria in tokamak fusion experiments and reactors. The approach involves representing the plasma current as a linear combination of basis functions using principal component analysis of plasma toroidal current densities (J t ) from the EFIT-AI equilibrium database. Then utilizing EFIT's Green's function tables, basis functions are created for the poloidal flux (ψ) and diagnostics generated from the toroidal current (J t ). Similar to the idea of a physics-informed neural network (NN), this physically enforces consistency between ψ, J t , and the synthetic diagnostics. First, the predictive capability of a least squares technique to minimize the error on the synthetic diagnostics is employed. The results show that the method achieves high accuracy in predicting ψ and moderate accuracy in predicting J t with median R 2 = 0.9993 and R 2 = 0.978, respectively. A comprehensive NN using a network architecture search is also employed to predict the coefficients of the basis functions. The NN demonstrates significantly better performance compared to the least squares method with median R 2 = 0.9997 and 0.9916 for J t and ψ, respectively. The robustness of the method is evaluated by handling missing or incorrect data through the least squares filling of missing data, which shows that the NN prediction remains strong even with a reduced number of diagnostics. Additionally, the method is tested on plasmas outside of the training range showing reasonable results.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2022

Decoding the protein–ligand interactions using parallel graph neural networks

Abstract Protein–ligand interactions (PLIs) are essential for biochemical functionality and their identification is crucial for estimating biophysical properties for rational therapeutic design. Currently, experimental characterization of these properties is the most accurate method, however, this is very time-consuming and labor-intensive. A number of computational methods have been developed in this context but most of the existing PLI prediction heavily depends on 2D protein sequence data. Here, we present a novel parallel graph neural network (GNN) to integrate knowledge representation and reasoning for PLI prediction to perform deep learning guided by expert knowledge and informed by 3D structural data. We develop two distinct GNN architectures: $$\hbox {GNN}_{\mathrm{F}}$$ GNN F is the base implementation that employs distinct featurization to enhance domain-awareness, while $$\hbox {GNN}_{\mathrm{P}}$$ GNN P is a novel implementation that can predict with no prior knowledge of the intermolecular interactions. The comprehensive evaluation demonstrated that GNN can successfully capture the binary interactions between ligand and protein’s 3D structure with 0.979 test accuracy for $$\hbox {GNN}_{\mathrm{F}}$$ GNN F and 0.958 for $$\hbox {GNN}_{\mathrm{P}}$$ GNN P for predicting activity of a protein–ligand complex. These models are further adapted for regression tasks to predict experimental binding affinities and $$\hbox {pIC}_{\mathrm{50}}$$ pIC 50 crucial for compound’s potency and efficacy. We achieve a Pearson correlation coefficient of 0.66 and 0.65 on experimental affinity and 0.50 and 0.51 on $$\hbox {pIC}_{\mathrm{50}}$$ pIC 50 with $$\hbox {GNN}_{\mathrm{F}}$$ GNN F and $$\hbox {GNN}_{\mathrm{P}}$$ GNN P , respectively, outperforming similar 2D sequence based models. Our method can serve as an interpretable and explainable artificial intelligence (AI) tool for predicted activity, potency, and biophysical properties of lead candidates. To this end, we show the utility of $$\hbox {GNN}_{\mathrm{P}}$$ GNN P on SARS-Cov-2 protein targets by screening a large compound library and comparing the prediction with the experimentally measured data.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

SAGIPS: a physics-inspired scalable asynchronous generative inverse-problem solver

Abstract Solving large-scale inverse problems using deep-learning algorithms have become an essential part of modern research and industrial applications. The complexity of the underlying inverse problem may require the utilization of high performance computing systems which poses a challenge on the algorithmic design of the inverse problem solver. Most deep learning algorithms require, due to their design, custom parallelization techniques in order to be resource efficient while showing a reasonable convergence. In this paper we introduce a S calable A synchronous G enerative I nverse P roblem S olver (SAGIPS) on high-performance computing systems. We present a workflow that utilizes an asynchronous ring-allreduce algorithm to transfer the gradients of the generator network across multiple GPUs. Experiments with a scientific proxy application demonstrate that SAGIPS shows near linear weak scaling, together with a convergence quality that is comparable to traditional methods. The approach presented here allows leveraging Generative Adverserial Network across multiple GPUs, promising advancements in solving complex inverse problems at scale.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Differential $t\overline{t}$ cross-section measurements using boosted top quarks in the all-hadronic final state with 139 fb -1 of ATLAS data

Measurements of single-, double-, and triple-differential cross-sections are presented for boosted top-quark pair-production in 13 TeV proton–proton collisions recorded by the ATLAS detector at the LHC. The top quarks are observed through their hadronic decay and reconstructed as large-radius jets with the leading jet having transverse momentum (pT) greater than 500 GeV. The observed data are unfolded to remove detector effects. The particle-level cross-section, multiplied by the $t\overline{t}$ $\rightarrow$ $WWb$$\overline{b}$ branching fraction and measured in a fiducial phase space defined by requiring the leading and second-leading jets to have p T > 500 GeV and p T > 350 GeV, respectively, is 331 ± 3(stat.) ± 39(syst.) fb. This is approximately 20% lower than the prediction of ${398}^{+48}_{-49}$ fb by POWHEG+PYTHIA 8 with next-to-leading-order (NLO) accuracy but consistent within the theoretical uncertainties. Results are also presented at the parton level, where the effects of top-quark decay, parton showering, and hadronization are removed such that they can be compared with fixed-order next-to-next-to-leading-order (NNLO) calculations. The parton-level cross-section, measured in a fiducial phase space similar to that at particle level, is 1.94 ± 0.02(stat.) ± 0.25(syst.) pb. This agrees with the NNLO prediction of ${1.96}^{+0.02}_{-0.17}$ pb. Reasonable agreement with the differential cross-sections is found for most NLO models, while the NNLO calculations are generally in better agreement with the data. The differential cross-sections are interpreted using a Standard Model effective field-theory formalism and limits are set on Wilson coefficients of several four-fermion operators.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2025

Functional Relevance of CASP16 Nucleic Acid Predictions as Evaluated by Structure Providers

ABSTRACT Accurate biomolecular structure prediction enables the prediction of mutational effects, the speculation of function based on predicted structural homology, the analysis of ligand binding modes, experimental model building, and many other applications. Such algorithms to predict essential functional and structural features remain out of reach for biomolecular complexes containing nucleic acids. Here, we report a quantitative and qualitative evaluation of nucleic acid structures for the CASP16 blind prediction challenge by 12 of the experimental groups who provided nucleic acid targets. Blind predictions accurately model secondary structure and some aspects of tertiary structure, including reasonable global folds for some complex RNAs; however, predictions often lack accuracy in the regions of highest functional importance. All models have inaccuracies in non‐canonical regions where, for example, the nucleic‐acid backbone bends, deviating from an A‐form helix geometry, or a base forms a non‐standard hydrogen bond (not a Watson‐Crick base pair). These bends and non‐canonical interactions are integral to forming functionally important regions such as RNA enzymatic active sites. Additionally, the modeling of conserved and functional interfaces between nucleic acids and ligands, proteins, or other nucleic acids remains poor. For some targets, the experimental structures may not represent the only structure the biomolecular complex occupies in solution or in its functional life cycle, posing a future challenge for the community.

Biochemistry & Molecular Biology↗
DOE OSTI2023

Rapid solidified ductile Cu-Al-Mn ribbon and its elastocaloric potential

Abstract Cu-Al-Mn alloys display martensitic transformation over a wide range of temperatures. In addition to low cost, this alloy is known for its low transformation stress with reasonable latent heat favoring elastocaloric applications. However, the ductility of Cu-Al-Mn can be limited owing to ordering and intergranular fracture. Through rapid solidification by melt spinning, we show that Cu-Al-Mn ribbon can be made highly ductile (greater than 8% tensile strain in the as-spun state and 10% tensile strain after heat treatment). The ductility of the melt-spun ribbon is related to the suppression of L2 1 ordering that is characterized through magnetic property measurement. Heat treatment of the ribbon promotes bamboo grain formation, and the latent heat is increased to 6.4 J g −1 . Under tensile conditions, we show that the ribbon exhibited about 4 °C temperature change (4.4 °C on heating and 4.2 °C on cooling from 6.3% strain).

36 MATERIALS SCIENCE↗
DOE OSTI2024

Fiducial and differential cross-section measurements of electroweak $W\gamma jj$ production in $pp$ collisions at $\sqrt{s} = 13$ TeV with the ATLAS detector

The observation of the electroweak production of a W boson and a photon in association with two jets, using pp collision data at the Large Hadron Collider at a centre of mass energy of $\sqrt{s} = 13$ TeV, is reported. The data were recorded by the ATLAS experiment from 2015 to 2018 and correspond to an integrated luminosity of 140 fb -1 . This process is sensitive to the quartic gauge boson couplings via the vector boson scattering mechanism and provides a stringent test of the electroweak sector of the Standard Model. Events are selected if they contain one electron or muon, missing transverse momentum, at least one photon, and two jets. Multivariate techniques are used to distinguish the electroweak $W\gamma jj$ process from irreducible background processes. The observed significance of the electroweak $W\gamma jj$ process is well above six standard deviations, compared to an expected significance of 6.3 standard deviations. Fiducial and differential cross sections are measured in a fiducial phase space close to the detector acceptance, which are in reasonable agreement with leading order Standard Model predictions from MADGRAPH5+PYTHIA8 and SHERPA. The results are used to constrain new physics effects in the context of an effective field theory.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

Observation of Significant Flavor-SU(3) Breaking in the Kaon Wave Function at 12 < Q 2 < 25 GeV 2 and Discovery of the Charmless Decay ψ ( 3770 ) → K S 0 K L 0

We present cross sections for the reaction e + e − → K S 0 K L 0 at center-of-mass energies ranging from 3.51 to 4.95 GeV using data samples collected in the BESIII experiment, corresponding to a total integrated luminosity of 26.5 fb − 1 . The ratio of neutral-to-charged kaon form factors at large momentum transfers ( 12 < Q 2 < 25 GeV 2 ) is determined to be 0.21 ± 0.01 , which indicates a small but significant effect of flavor-SU(3) breaking in the kaon wave function, and, consequently, excludes the possibility that flavor-SU(3) breaking is the primary reason for the strong experimental violation of the pQCD prediction | F ( π ± ) | / | F ( K ± ) | = f π 2 / f K 2 , where F ( π ± ) and F ( K ± ) are the form factors, and f π and f K are the decay constants of charged pions and kaons, respectively. We also observe a significant signal for the charmless decay ψ ( 3770 ) → K S 0 K L 0 for the first time. Within a 1 σ contour of the likelihood value, the branching fraction for ψ ( 3770 ) → K S 0 K L 0 is determined to be B = ( 2.6 3 − 1.59 + 1.40 ) × 10 − 5 , and the relative phase between the continuum and ψ ( 3770 ) amplitudes is ϕ = ( − 0.3 9 − 0.10 + 0.05 ) π . The branching fraction is in good agreement with the S - and D -wave charmonia mixing scheme proposed in the interpretation of the “ ρ π puzzle” between J / ψ and ψ ( 3686 ) decays. Published by the American Physical Society 2024

Physics↗
DOE OSTI2025

Recent Developments in DFTB+, a Software Package for Efficient Atomistic Quantum Mechanical Simulations

DFTB+ is a flexible, open-source software package developed by its community, designed for fast and efficient atomistic quantum mechanical simulations. It employs various methods that approximate density functional theory (DFT), such as density functional-based tight binding (DFTB) and the extended tight binding (xTB) approach allowing simulations of large systems over extended time scales with reasonable accuracy, while being significantly faster than traditional ab initio methods. In recent years, several new extensions of the DFTB method have been developed and implemented in the DFTB+ program package in order to improve the accuracy and generality of the available simulation results. In this paper, we review those enhancements, show several use case examples and discuss the strengths and limitations of its features.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Illuminating the Material World: Autonomous Microscopy to Understand Order, Disorder, and Everything In Between

Artificial intelligence (AI) holds immense promise for revolutionizing microscopy, yet its widespread adoption has been hindered by challenges ranging from user inexperience to limited model transferability and difficulties in operationalizing machine learning. This presentation showcases our approach to developing practical autonomy for materials discovery, aiming to accelerate the integration of AI into everyday microscopy workflows. As shown in Fig. 1, I will focus on three key areas: understanding order-disorder transitions, quantifying point defects, and achieving truly device-scale microscopy. First, I will demonstrate the power of multi-modal knowledge graphs for integrating diverse microscopy data. By combining imaging, spectroscopy, and diffraction data, these graphs provide a holistic view of material behavior, capturing the intricate relationships between different modalities [1,2]. I will present a case study on how these models illuminate the structural and chemical changes associated with irradiation in oxide thin films, revealing critical insights for designing materials for extreme environments like spaceflight and nuclear energy. Specifically, I will show how multi-modal analysis clarifies the evolution of order-disorder transitions under irradiation, a key factor influencing material performance in these applications. Next, I will address the challenge of quantifying point defects in 2D materials. We demonstrate the application of computer vision and transfer learning to accurately identify and classify various defect types, such as vacancies and substitutional atoms, and to quantify their concentrations. This information is crucial for understanding and tailoring the properties of 2D materials for applications in electronics, optoelectronics, and catalysis. For example, I will show how our models can characterize the topological distribution of point defects in MXene transition metal carbides, providing valuable insights for optimizing their performance in energy storage and separation science. Finally, I will discuss our progress toward autonomous device-scale microscopy [3,4]. We are fundamentally redesigning electron microscopes around the principles of machine reasoning, enabling automation beyond basic tasks like sample navigation and data acquisition to include sophisticated experimental design. This approach paves the way for truly reproducible and massively scaled analysis campaigns. I will emphasize the importance of autonomous microscopy platforms for high-throughput materials discovery and characterization, facilitating the rapid screening of materials for a broad range of applications and accelerating the development of next-generation technologies.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Modernizing the Nuclear Industry and New Ways of Working

Nuclear energy is recognized as the most feasible energy source towards achieving nation-wide net-zero goals (COP 208, MIT). Additionally, the demand for reliable, consistent energy supply is soaring with the construction of energy hungry data centers and AI tools. This means the construction of new nuclear and the continuation of current nuclear are a high national priority. However, this leaves the nuclear industry with a workforce challenge. The rising demand for nuclear power is threatened by an underpopulated workforce. There exist a few contributors to the decreased workforce such as skewed workforce demographics resulting in large-scale workforce retirement. Also, some plants have observed difficulty hiring and retaining skill sets that typically comprise the nuclear power workforce – specifically skilled craftsman. Those who are graduating with desirable skills are seeking employment in sectors that are more modern and culturally more aligned with younger generational values. Lastly, within the nuclear sector, skilled employees will be a competitive commodity as advanced plants come online offering work environments that incorporate modern technologies, skill sets, and opportunities for career advancement. These factors emphasize the need for legacy plants to modernize with a focus on the workplace environment and culture that meets younger generations’ skills, cultural expectations, and desire for advancement opportunities. Nuclear plants are actively engaged in deploying advancements that improve the management of systems, structures, and components as well as process improvements and technologies that can reduce the costs of operation and maintenance. However, the advancements must also be analyzed, reviewed and deployed in a manner that considers the impact on how people within the organization collaborate, communicate, make decisions, and solve problems. Neglecting to consider the cultural impact of modernization risks poor adoption, unrealized opportunities, or insignificant change toward helping attract new employees. The traditional way of working is not necessarily the way new generations want to engage with their employer. For instance, in a survey performed by North American Young Generation in Nuclear the top three reasons for younger nuclear employees to seek other job opportunities was seeking better work-life balance, a lack of advancement opportunities, and work culture and leadership style differences (Smyth et al. 2022). Integrated Operations for Nuclear (ION) is an approach for the nuclear industry to create long-term strategic modernization plans and analyze each advancement for the impact on people and processes and measure how those impacts flow up to support high-level, long-term plant goals and requirements. One principle of ION is to replace the current labor-centric operation style in legacy nuclear plants with data-centric collaborative operation styles. This paper will explain how transitioning operations to leverage centralized skills and responsibility, multi-skilled teams, and collaborative decision making can flatten the hierarchical structure in plants, enable greater individual efficacy while also offering more experiences and advancement opportunities to staff. Adopting the ION way of working shifts the industry mind-set towards more networked, collaborative, and modern way of working that will attract skilled, next-generation workers to operate the legacy nuclear fleet.

99 - GENERAL AND MISCELLANEOUS↗
DOE OSTI2025

Knowledge Oriented Graph Unified Transformer (KOGUT) v0.1

KOGUT — Knowledge Oriented Graph Unified Transformer KOGUT implements the Relational Graph Transformer (RelGT) architecture for knowledge graph link prediction in biological domains, with a primary focus on microbial growth media prediction. While the original RelGT (arXiv:2505.10960) targets relational tables, time series, and multi-table databases, KOGUT adapts this architecture for heterogeneous biological knowledge graphs, providing first-in-class AI predictive models for microbial cultivation. Key Adaptations Beyond Original RelGT: - Knowledge Graph Focus: Applied to biological KGs with semantic node types (taxa, chemicals, media, phenotypes, environments) versus generic relational database tables, trained on the KG-Microbe knowledge graph (1.3M entities, 2.9M edges, 24 relation types). - Multimodal Node Encoding: Integrates node labels, categories, descriptions, and synonyms from KG metadata through learned embedding layers—adapting relational column features to graph node attributes with textual semantics. - Extended K-Hop Subgraph Strategy: Optimized neighborhood sampling (3-hop default, configurable up to 200 nodes) tuned for sparse biological networks, building on the original local-global attention framework with biological relation preservation. - Biolink Predicate Preservation: Type-specific transformations for 24 biological edge semantics (occurs_in, consumes, produces, has_phenotype, subclass_of) beyond standard relational foreign keys, enabling multi-relation link prediction. - Inductive Learning Support: Enables zero-shot predictions for novel taxa through feature-based embeddings (temperature, oxygen requirements, gram stain, cell shape), extending the original transductive relational benchmark scope to uncultured microorganisms. CheapSOTA Performance Optimizations (This Distribution): - VQ-EMA Centroid Attention: Vector quantization with exponential moving average for improved global context modeling (+5-10% MRR improvement). - HDF5 Precomputed Data Loading: One-time preprocessing of k-hop subgraphs to eliminate redundant graph traversals (2-5× training speedup). - Distributed Data Parallel Training: Multi-GPU support for scaling to larger knowledge graphs (tested on 4× NVIDIA A100 GPUs at NERSC Perlmutter). - Mixed Precision Training: Automatic mixed precision (AMP) for memory efficiency and faster training. Advantages Over Standard Knowledge Graph Embedding Models: Combines RelGT's proven multi-element tokenization (features, type, hop, structure) with graph-native biological representations, enabling interpretable link prediction across heterogeneous entities that standard embedding models (TransE, RotatE, ComplEx) and table-based transformers cannot directly model. Achieves near-perfect performance on microbial growth media prediction (MRR: 0.9966, Precision@1: 0.9932, Hit@10: 1.0000) while maintaining explainability through attention-based reasoning over biological pathways. Training Data: - KG-Microbe merged knowledge graph: 1,379,337 nodes, 2,960,472 edges - 24 biological relation types including taxonomic hierarchies, metabolic interactions, phenotype associations, and environmental relationships - Primary prediction task: Growth media suitability for microbial taxa (biolink:occurs_in, 50K edges) - Multi-relation capability: Predicts links for any of the 24 relation types, including chemical consumption/production, phenotype associations, and taxonomic classification Citation: Original RelGT Architecture: Dwivedi et al., "Relational Graph Transformer", arXiv:2505.10960, 2025 KOGUT Implementation: Knowledge Oriented Graph Unified Transformer for Microbial Growth Media Prediction Developed at Lawrence Berkeley National Laboratory (LBNL) Trained on NERSC Perlmutter supercomputer

Joachimiak, Marcin [Lawrence Berkeley National Lab↗
DOE OSTI2026

Holistic energy analysis method for thermal management architectures of data centers

Modern high-performance computing (HPC) data centers (DCs), particularly those supporting energy-intensive artificial intelligence (AI) workloads, face escalating thermal management challenges that degrade performance through thermal throttling and drive up cooling power consumption and operational costs. To address this challenge, many have developed a wide variety of thermal management solutions (single-phase, two-phase, direct, indirect, hybrid, and more) which attempt to cool HPC DCs effectively while attempting to minimize overall system power consumption. However, the analysis of these solutions and methods to effectively compare one with another is lacking. Overall power usage effectiveness (PUE) and total-power usage effectiveness (TUE) provide a metric to quantify power consumption but fail to identify components in the system which require further optimization. To address this, we propose a holistic analytical framework – the waterfall diagram (WFD) – which leverages a waterfall chart methodology, offering a comprehensive visualization of both the thermal management system loop and heat flow pathways from individual server components to the outdoor ambient. Use of the WFD enables graphical estimations of power efficiency and cooling performance across each component of a DC cooling system and complements Sankey-style energy flow visualizations by additionally resolving stage-wise temperature changes and incremental TUE contributions. The framework is used in conjunction with simulation-based approaches, to conduct a detailed pressure drop and flow distribution analysis aimed at identifying the optimal coolant distribution architecture for a single-phase direct-to-chip water-cooled DC, which serves as the baseline for subsequent WFD analysis. Among the evaluated architectures, the 3 U modular coolant distribution architecture is found to demonstrate the best performance, considering minimal pressure drop and uniform flow distribution. In addition, TUE is calculated for each cooling loop component based on its associated pressure drop and corresponding pumping power, which are integrated into the WFD. This correlation between TUE and local temperature offers immediate insight into the power efficiency and thermal performance contributions of individual components, facilitating further development and optimization. Examples of WFD applications are presented under varying thermal loads and ambient conditions, demonstrating reasonable cooling strategies. Notably, the 3 U modular architecture maintains a consistent chip case temperature of 85°C, achieving a TUE of 1.016 at ambient temperature of 47°C, and a TUE of 1.026 at ambient temperature of 52°C. The WFD methodology provides an efficient, holistic, and streamlined framework for DC thermal management architecture assessment and enables design optimization which is important for addressing the thermal-fluidic energy challenges of current and next-generation DCs.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2025

Discovery of a z ∼ 0.8 ultra steep spectrum radio halo in the MeerKAT-South Pole Telescope Survey

Radio haloes are diffuse synchrotron sources that trace the turbulent intracluster medium (ICM) of galaxy clusters. However, their origin remains unknown. Two main formation models have been proposed: the hadronic model, in which relativistic electrons are continuously injected by cosmic-ray protons; and the leptonic turbulent re-acceleration model, where cluster mergers re-energize electrons in situ. A key discriminant between the two models would be the existence of ultra-steep spectrum radio haloes (USSRHs), which can only be produced through turbulent re-acceleration. Here, we report the discovery of an USSRH in the galaxy cluster SPT-CLJ2337–5942 at redshift $z = 0.78$ in the MeerKAT-South Pole Telescope 100 deg$^2$ UHF (0.58–1.09 GHz) survey. This discovery is noteworthy for two primary reasons: it is the highest redshift USSRH system to date; and the close correspondence of the radio emission with the thermal ICM as traced by Chandra X-ray observations, further supporting the leptonic re-acceleration model. The halo is underluminous for its mass, consistent with a minor merger origin, which produces steep-spectrum, lower luminosity haloes. This result demonstrates the power of wide-field, high-fidelity, low-frequency ($\lesssim 1$ GHz) surveys like the MeerKAT-SPT 100 deg$^2$ programme to probe the origin and evolution of radio haloes over cosmic time, ahead of the Square Kilometre Array.

X-rays: galaxies↗