Search NASA⌕ Search

SEARCH · Search NASA

Results for “flow matching”

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 19 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

Efficient many-jet event generation with flow matching

We apply for the first time, to the best of our knowledge, the flow matching method to the problem of phase-space sampling for event generation in high-energy collider physics. By training the model to remap the random numbers used to generate the momenta and helicities of the scattering matrix elements as implemented in the portable partonic event generator pepper, we find substantial efficiency improvements in the studied processes. We focus our study on the highest final-state multiplicities in Drell-Yan and top-antitop pair production used in simulated samples for the Large Hadron Collider, which computationally are the most relevant ones. We find that the unweighting efficiencies improve by factors of 184 and 25, respectively, when compared to the standard approach of using a vegas-based optimization. We also compare continuous normalizing flows trained with flow matching against the previously studied normalizing flows based on coupling layers and find that the former leads to better results, faster training and a better scaling behavior across the studied multiplicity range, while the latter evaluate faster. When combining the advantages of both methods using the regflow approach, we find parton-level unweighted event generation walltime gains of about a factor of 10 at the highest final-state multiplicities.

Bothmann, E. [CERN; Gottingen U.] (ORCID:000000016↗
DOE OSTI2026

Flow matching meets biology and life science: a survey

Over the past decade, advances in generative modeling, such as generative adversarial networks, masked autoencoders, and diffusion models, have significantly transformed biological research and discovery, enabling breakthroughs in molecule design, protein generation, catalysis discovery, drug discovery, and beyond. At the same time, biological applications have served as valuable testbeds for evaluating the capabilities of generative models. Recently, flow matching has emerged as a powerful and efficient alternative to diffusion-based generative modeling, with growing interest in its application to problems in biology and life sciences. This paper presents the first comprehensive survey of recent developments in flow matching and its applications in biological domains. We begin by systematically reviewing the foundations and variants of flow matching, and then categorize its applications into three major areas: biological sequence modeling, molecule generation and design, and peptide and protein generation. For each, we provide an in-depth review of recent progress. We also summarize commonly used datasets and software tools, and conclude with a discussion of potential future directions.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Flow matching beyond kinematics: Generating jets with particle identification and trajectory displacement information

We introduce the first generative model trained on the etlass dataset. Our model generates jets at the constituent level, and it is a permutation-equivariant continuous normalizing flow (CNF) trained with the flow matching technique. It is conditioned on the jet type, so that a single model can be used to generate the ten different jet types of etlass. For the first time, we also introduce a generative model that goes beyond the kinematic features of jet constituents. The etlass dataset includes more features, such as particle-ID and track impact parameter, and we demonstrate that our CNF can accurately model all of these additional features as well. Our generative model for etlass expands on the versatility of existing jet generation techniques, enhancing their potential utility in high-energy physics research, and offering a more comprehensive understanding of the generated jets. Published by the American Physical Society 2025

Birk, Joschka (ORCID:0000000219310127)↗
DOE OSTI2025

A comparison of probabilistic generative frameworks for molecular simulations

Generative artificial intelligence is now a widely used tool in molecular science. Despite the popularity of probabilistic generative models, numerical experiments benchmarking their performance on molecular data are lacking. Here, in this work, we introduce and explain several classes of generative models, broadly sorted into two categories: flow-based models and diffusion models. We select three representative models: neural spline flows, conditional flow matching, and denoising diffusion probabilistic models, and examine their accuracy, computational cost, and generation speed across datasets with tunable dimensionality, complexity, and modal asymmetry. Our findings are varied, with no one framework being the best for all purposes. In a nutshell, (i) neural spline flows do best at capturing mode asymmetry present in low-dimensional data, (ii) conditional flow matching outperforms other models for high-dimensional data with low complexity, and (iii) denoising diffusion probabilistic models appear the best for low-dimensional data with high complexity. Our datasets include a Gaussian mixture model and the dihedral torsion angle distribution of the Aib9 peptide, generated via a molecular dynamics simulation. We hope our taxonomy of probabilistic generative frameworks and numerical results may guide model selection for a wide range of molecular tasks.

Artificial intelligence↗
DOE OSTI2026

Monte Carlo Event Generation with Continuous Normalizing Flows

We apply continuous normalizing flows trained with the flow matching method to the problem of phase-space sampling in Monte Carlo event generation for high-energy collider physics. Focusing on lepton-pair and top-quark pair production with multiple jets, the two computationally most expensive processes at the Large Hadron Collider, we train helicity-conditioned continuous normalizing flows to remap the random numbers used in matrix element evaluation. Compared to standard methods, we achieve unweighting efficiency improvements by factors of up to 184 and 25 for the two processes at their respective highest jet number, at the cost of an increased evaluation time. When combining the advantages of continuous normalizing flows with the fast evaluation times of coupling-layer-based flows, using the RegFlow approach, we find parton-level unweighted event generation walltime gains of about a factor of 10 at the highest jet numbers. These substantial gains highlight the promise of samplers based on machine learning for next-generation collider experiments.

Bothmann, Enrico [CERN; Gottingen U.] (ORCID:00000↗
DOE OSTI2025

Universal energy-speed-accuracy trade-offs in driven nonequilibrium systems

The connection between measure theoretic optimal transport and dissipative nonequilibrium dynamics provides a language for quantifying nonequilibrium control costs, leading to a collection of thermodynamic speed limits, which rely on the assumption that the target probability distribution is perfectly realized. This is almost never the case in experiments or numerical simulations, so here we address the situation in which the external controller is imperfect. We obtain a lower bound for the dissipated work in generic nonequilibrium control problems that (1) is asymptotically tight and (2) matches the thermodynamic speed limit in the case of optimal driving. Along with analytically solvable examples, we refine this imperfect driving notion to systems in which the controlled degrees of freedom are slow relative to the nonequilibrium relaxation rate, and identify independent energy contributions from fast and slow degrees of freedom. Furthermore, we develop a strategy for optimizing minimally dissipative protocols based on optimal transport flow matching, a generative machine learning technique. Furthermore, this latter approach ensures the scalability of both the theoretical and computational framework we put forth. Crucially, we demonstrate that we can compute the terms in our bound numerically using efficient algorithms from the computational optimal transport literature and that the protocols we learn saturate the bound.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2026

CaloTrilogy: Toward a Breakthrough in One-Step, End-to-End, Physics-Guided Shower Generation for Modern Calorimeters

High-precision calorimeter simulation at current and future colliders imposes rapidly growing computational demands, motivating the development of machine-learning surrogates for traditional Monte Carlo tools such as Geant4. Flow matching and diffusion-based generative models have become leading approaches for high-dimensional fast simulation because of their sample quality, but typically require ${\cal O}(100)$ function evaluations at inference and often rely on auxiliary networks to constrain global observables, compromising streamlined end-to-end generation. We introduce a unified framework that improves the balance between speed, shower quality, and physics fidelity. The method combines: (i) an average velocity field integrator that enables sampling in one or a few evaluations; (ii) a learned generative prior in shower space, constructed from data rather than random noise; and (iii) physics-guided loss terms that impose inductive biases on key observables during training. These elements are training time regularizers, preserving end-to-end inference with no additional cost. With only one or a few evaluation steps, the model achieves shower quality competitive with state-of-the-art flow and diffusion approaches, tested on several public high granularity calorimeter datasets. The results demonstrate inter-layer shower structure consistent with the underlying physics, providing a strong candidate for future fast simulation workflows.

Jiang, Cheng [Edinburgh U.]↗
DOE OSTI2025

CaloChallenge 2022: a community challenge for fast calorimeter simulation

Here, we present the results of the ‘Fast Calorimeter Simulation Challenge 2022’—the CaloChallenge. We study state-of-the-art generative models on four calorimeter shower datasets of increasing dimensionality, ranging from a few hundred voxels to a few tens of thousand voxels. The 31 individual submissions span a wide range of current popular generative architectures, including variational autoencoders (VAEs), generative adversarial networks (GANs), normalizing flows, diffusion models, and models based on conditional flow matching. We compare all submissions in terms of quality of generated calorimeter showers, as well as shower generation time and model size. To assess the quality we use a broad range of different metrics including differences in one-dimensional histograms of observables, KPD/FPD scores, AUCs of binary classifiers, and the log-posterior of a multiclass classifier. The results of the CaloChallenge provide the most complete and comprehensive survey of cutting-edge approaches to calorimeter fast simulation to date. In addition, our work provides a uniquely detailed perspective on the important problem of how to evaluate generative models. As such, the results presented here should be applicable for other domains that use generative AI and require fast and faithful generation of samples in a large phase space.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2026

Generative models on phase space

Deep generative models such as diffusion and flow matching are powerful machine learning tools capable of learning and sampling from high-dimensional distributions. They are particularly useful when the training data appears to be concentrated on a submanifold of the data embedding space. For high-energy physics data, consisting of collections of relativistic energy-momentum 4-vectors, this submanifold can enforce extremely strong physically-motivated priors, such as energy and momentum conservation. If these constraints are learned only approximately, rather than exactly, this can inhibit the interpretability and reliability of such generative models. To remedy this deficiency, we introduce generative models which are, by construction, confined at every step of their sampling trajectory to the manifold of massless N-particle Lorentz-invariant phase space in the center-of-momentum frame. In the case of diffusion models, the "pure noise" forward process endpoint corresponds to the uniform distribution on phase space, which provides a clear starting point from which to identify how correlations among the particles emerge during the reverse (de-noising) process. We demonstrate that our models are able to learn both few-particle and many-particle distributions with various singularity structures, paving the way for future interpretability studies using generative models trained on simulated jet data.

Bogorad, Zachary [Fermilab]↗
DOE OSTI2025

Self-regulating behavior of hybrid membrane systems as demonstrated in an element-scale forward osmosis-reverse osmosis hybrid system

Hybrid membrane systems can be difficult to design due to the requisite flow rate matching between up- and downstream unit operations. In this work, we use a forward osmosis-reverse osmosis (FO-RO) hybrid system to demonstrate how some membrane systems can exhibit self-regulating behavior due to osmotic coupling. This can reduce the need for complex control systems for flow balancing. We show this behavior using a module-scale test bed that can mimic the behavior of larger scale operations. The system shows permeate flow rate near-convergence between the FO and RO modules after startup or when perturbed by a change in RO module pressure. The behavior of this hybrid system demonstrates that some membrane operations can exploit osmotic interdependence, rather than expensive control systems, to achieve steady state operation.

Debottlenecking↗
DOE OSTI2026

FunDiff: diffusion models over function spaces for physics-informed generative modeling

Recent advances in generative modeling-particularly diffusion models and flow matching-have been widely used for synthesizing discrete data such as images and videos. However, adapting these models to physical applications remains challenging, as the quantities of interest are continuous functions governed by complex physical laws. To address this, we introduce FunDiff, an efficient and robust framework for generative modeling in function spaces. FunDiff combines a latent diffusion process with a function autoencoder architecture to handle input functions with varying discretizations, generates continuous functions that can be evaluated at arbitrary locations, and seamlessly incorporate physical priors. These priors are enforced through architectural constraints or physics-informed loss functions, ensuring that generated samples satisfy fundamental physical laws. We theoretically establish minimax optimality guarantees for density estimation in function spaces, demonstrating that diffusion-based estimators achieve optimal convergence rates under suitable regularity conditions. We further demonstrate the practical effectiveness of FunDiff across diverse applications in fluid dynamics and solid mechanics. Empirical results indicate that our method can generate physically consistent samples with high fidelity to the target distribution, and exhibit robustness to noisy and low-resolution data.

Wang, Sifan [Yale University, New Haven, CT (Unite↗
DOE OSTI2025

Nonadiabatic Force Matching for Alchemical Free-Energy Estimation

We propose a method to compute free-energy differences from nonadiabatic alchemical transformations by using flow-based generative models. The method, nonadiabatic force matching, hinges on estimating the dissipation along an alchemical switching process in terms of a nonadiabatic force field that can be learned through stochastic flow matching. The learned field can be used in conjunction with short-time trajectory data to evaluate upper and lower bounds on the alchemical free energy that variationally converge to the exact value if the field is optimal. Applying the method to evaluate the alchemical free energy of atomistic models shows that it can substantially reduce the simulation cost of a free-energy estimate at a negligible loss of accuracy when compared with thermodynamic integration.

Computational chemistry↗
DOE OSTI2026

leblanc-lab/MA-PFN

This repository is a compact, runnable implementation of the Metric-Aware Particle Flow Network (MA-PFN) used to regress the Energy Mover's Distance (EMD). It trains MA-PFN alongside a matched, unconstrained Particle Flow Network (PFN), then compares their held-out accuracy and metric properties.

LeBlanc, Matt [@BrownUniversity]↗
DOE OSTI2026

Understanding spectral dependance of laser-induced damage precursors in dielectric materials (Abbreviated report_23-ERD-006)

High peak and average power laser systems are typically limited by the handling fluence of the optical components. In particular, the multilayer dielectric coatings are known to be much lower operational fluence than the more ideal bulk materials. In this project we proposed and succeeded in probing different established classes of damage prone precursors at different wavelengths to understand their fundamental laser damage response as a function of wavelength. This study helped to shed light on the fundamental physics of the non-linear precursors that govern laser damage phenomena for ns-regime pulsed laser damage. In this study, we utilized the onsite coating capabilities (VPL, IBS coating lab) to purposefully generate laser damage-prone precursors in hafnia-based coatings (both single and multi-layer coatings). Specifically, we engineered coatings with craze lines initiated by nodules, generated coatings with our xenon-based coating process to suppress nanobubble formation and generated hafnia coatings under controlled oxygen flow conditions to study hafnia sub-oxides and oxygen flow dependance. In these studies we found that craze lines are rife with precursors that are sensitive to ultra-violet light but not to infrared light; we found that the removal of nanobubbles helps with all wavelengths tested, but is most impactful for ultra-violet light; we also found that ultra-violet laser damage performance of hafnia is closely matched to oxygen flow rate, while the infrared performance may be slightly better at lower flow rates. During this LDRD we also successfully stood up a new laser damage testing capability, namely a wavelength agile damage test station to study the spectral response of known laser-induced damage precursors. This is a unique and important capability for Lawrence Livermore National Lab, allowing us to understand the spectral response of materials under high intensity irradiation and damage. This is a vital tool to understand non-linear optical response at wavelengths that we have previously been unable to test at.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Acceleration of Power System Dynamic Simulations Using a Deep Equilibrium Layer and Neural ODE Surrogate

The dominant paradigm for power system dynamic simulation is to build system-level simulations by combining physics-based models of individual components. The sheer size of the system along with the rapid integration of inverter-based resources exacerbates the computational burden of running time domain simulations. Here, in this paper, we propose a data-driven surrogate model based on implicit machine learningspecifically deep equilibrium layers and neural ordinary differential equationsto learn a reduced order model of a portion of the full underlying system. The data-driven surrogate achieves similar accuracy and reduction in simulation time compared to a physics-based surrogate, without the constraint of requiring detailed knowledge of the underlying dynamic models. This work also establishes key requirements needed to integrate the surrogate into existing simulation workflows; the proposed surrogate is initialized to a steady state operating point that matches the power flow solution by design.

Neural ordinary differential equations↗
DOE OSTI2026

Heat Transfer Experiments of a 1st Stage Blade Cascade for Supercritical CO2 Oxy-Combustion Turbine Application

The results of internally cooled 1st stage blade (S1B) cascade testing in a supercritical CO2 environment is presented. The turbine blade design has been previously established for the end application of an oxy-combustion turbine operating in the Allam-Fetvedt cycle with turbine inlet conditions of 305 bar and 1150°C. The internally cooled blade features leading edge (LE) region impingement cooling, mid-section ribbed serpentine passages, and a pin-finned trailing edge (TE) region before cooling ejection holes. The geometry for the tested blade cascade has a cooled central blade with un-cooled blades on either side to match flowpath areas of the actual turbine. The flowpath reuses internal components previously employed for mid-section region ribbed serpentine passage experiments that established Nusselt number enhancement ratios over a range of Reynolds numbers from 100,000-400,000. New components include flow conditioning plates upstream and downstream of the blade cascade to adequately represent the flow field and blade external heat transfer coefficient profiles for the actual turbine. The cooled central blade utilizes uniform crystal temperature sensors (UCTS) with six sensors each on the blade pressure and suction surfaces distributed radially and from LE to TE. The post-processed UCTS quantified the maximum wall temperature seen at each installed sensor location. The test procedure consisted of establishing supercritical CO2 cooling flow temperature and flow rate and maintaining it throughout the test. The flow rate aims to match that for the actual in-service turbine blade design and is maintained through an orifice restriction to keep the pressure differential between internal cooling flow and external hot flow nearly constant. For the sCO2 flow path external to the blade, temperatures were ramped throughout the test via control of the test loop’s natural gas burner heater. The maximum temperature seen was 468°C and held constant for a duration of 10 minutes at which the blade metal temperature was predicted to be at its maximum before ramping down. For the turbine blade design for service inlet conditions, external flow path computational fluid dynamics (CFD) results and an internal cooling 1-D thermal and hydraulic flow network model using experimentally validated correlations served as thermal finite element (FE) boundary conditions to predict blade metal temperatures. These predicted temperatures were subsequently utilized in a structural FE model to predict blade life ratings dictated by Haynes 282 creep strength data, having a strong dependence on temperature. The boundary conditions experienced during testing are used in the same workflow and compared to the experimental results, with the goal of validating the analysis methodology and providing insight on the uncertainty in local metal temperature predictions.

20 FOSSIL-FUELED POWER PLANTS↗
DOE OSTI2025

Experimental Validation of Exact Burst Pressure Solutions for Thick-Walled Cylindrical Pressure Vessels

Burst pressure is one of the critical strength parameters used in the design and operation of pressure vessels because it represents the maximum pressure that a vessel can withstand before failing. Historically, the Barlow formula was used as a design base for estimating burst pressure. However, it does not consider the plastic flow response for ductile steels and is applicable only to thin-walled cylinders (i.e., the diameter to thickness ratio D/t ≥ 20). A new multiaxial plastic yield theory was developed to consider the plastic flow response, and the associated theoretical (i.e., Zhu–Leis) solution of burst pressure was obtained and has gained extensive applications in the pipeline industry because it was validated by different full-scale burst test datasets for large-diameter, thin-walled pipelines in a variety of steel grades from Grade B to X120. The Zhu–Leis flow theory of plasticity was recently extended to thick-walled pressure vessels, and the associated exact flow solution of burst pressure was obtained and is applicable to both thin and thick-walled cylindrical shells. Many full-scale burst tests are available for thin-walled line pipes in the pipeline industry, but limited pressure burst tests exist for thick-walled vessels. To validate the newly developed exact solutions of burst pressure for thick-walled cylinders, this paper conducts a series of burst pressure tests on small-diameter, thick-walled pipes. In particular, six burst tests are carried out for three thick-walled pipes in Grade B carbon steel. These pipes have a nominal diameter of 2.375 inches (60.33 mm) and three nominal wall thicknesses of 0.154, 0.218, and 0.344 inches (3.91, 5.54, and 8.74 mm), leading to D/t = 15.4, 10.9, and 6.9, respectively. With the burst test data, comparisons show that the Zhu–Leis flow solution of burst pressure matches well the burst test data for thick-walled pipes. Thus, these burst tests validate the accuracy of the Zhu–Leis flow solution of burst pressure for thick-walled cylindrical vessels.

42 ENGINEERING↗