Engineering Papers⌕ Search

Engineering topics

Huang, Yi

Publications and source records attributed to Huang, Yi.

At least 19 records

FM4NPP: A Scaling Foundation Model for Nuclear and Particle Physics

Large language models have revolutionized artificial intelligence by enabling large, generalizable models trained through self-supervision. This paradigm has inspired the development of scientific foundation models (FMs). However, applying this capability to experimental particle physics is challenging due to the sparse, spatially distributed nature of detector data, which differs dramatically from natural language. This work addresses if an FM for particle physics can scale and generalize across diverse tasks. We introduce a new dataset with more than 11 million particle collision events and a suite of downstream tasks and labeled data for evaluation. We propose a novel self-supervised training method for detector data and demonstrate its neural scalability with models that feature up to 188 million parameters. With frozen weights and task-specific adapters, this FM consistently outperforms baseline models across all downstream tasks. The performance also exhibits robust data-efficient adaptation. Further analysis reveals that the representations extracted by the FM are task-agnostic but can be specialized via a single linear mapping for different downstream tasks.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Unpaired image translation to mitigate domain shift in liquid argon time projection chamber detector responses

Deep learning algorithms often are developed and trained on a training dataset and deployed on test datasets. Any systematic difference between the training and a test dataset may severely degrade the final algorithm performance on the test dataset—what is known as the domain shift problem . This issue is prevalent in many scientific domains where algorithms are trained on simulated data but applied to real-world datasets. Typically, the domain shift problem is solved through various domain adaptation (DA) methods. However, these methods are often tailored for a specific downstream task, such as classification or semantic segmentation, and may not easily generalize to different tasks. This work explores the feasibility of using an alternative way to solve the domain shift problem that is not specific to any downstream algorithm. The proposed approach relies on modern Unpaired Image-to-Image (UI2I) translation techniques, designed to find translations between different image domains in a fully unsupervised fashion. In this study, the approach is applied to a domain shift problem commonly encountered in Liquid Argon Time Projection Chamber (LArTPC) detector research when seeking a way to translate samples between two differently distributed LArTPC detector datasets deterministically. This translation allows for mapping real-world data into the simulated data domain where the downstream algorithms can be run with much less domain-shift-related performance degradation. Conversely, using the translation from the simulated data to a real-world domain can increase the realism of the simulated dataset and reduce the magnitude of any systematic uncertainties. To evaluate the quality of the translations, we use both pixel-wise metrics and a downstream task to measure the effectiveness of UI2I methods for mitigating the domain shift problem. We adapted several popular UI2I translation algorithms to work on scientific data and demonstrated the viability of these techniques for solving the domain shift problem with LArTPC detector data. To facilitate further development of DA techniques for scientific datasets, the ‘Simple Liquid-Argon Track Samples’ dataset used in this study is also published.

97 MATHEMATICS AND COMPUTING↗

Effectiveness of denoising diffusion probabilistic models for fast and high-fidelity whole-event simulation in high-energy heavy-ion experiments

Artificial intelligence (AI) generative models, such as generative adversarial networks (GANs), variational autoencoders, and normalizing flows, have been widely used and studied as efficient alternatives for traditional scientific simulations. However, they have several drawbacks, including training instability and inability to cover the entire data distribution, especially for regions where data are rare. This is particularly challenging for whole-event, full-detector simulations in high-energy heavy-ion experiments, such as sPHENIX at the Relativistic Heavy Ion Collider and Large Hadron Collider experiments, where thousands of particles are produced per event and interact with the detector. This work investigates the effectiveness of denoising diffusion probabilistic models (DDPMs) as an AI-based generative surrogate model for the sPHENIX experiment that includes the heavy-ion event generation and response of the entire calorimeter stack. DDPM performance in sPHENIX simulation data is compared with a popular rival, GANs. Results show that both DDPMs and GANs can reproduce the data distribution where the examples are abundant (low-to-medium calorimeter energies). Nonetheless, DDPMs significantly outperform GANs, especially in high-energy regions where data are rare. Additionally, DDPMs exhibit superior stability compared to GANs. The results are consistent between both central and peripheral centrality heavy-ion collision events. Moreover, DDPMs offer a substantial speedup of approximately a factor of 100 compared to the traditional Geant4 simulation method.

42 ENGINEERING↗

Fast 2D Bicephalous Convolutional Autoencoder for Compressing 3D Time Projection Chamber Data

High-energy large-scale particle colliders produce data at high speed in the order of 1 terabytes per second in nuclear physics and petabytes per second in high energy physics. Developing real-time data compression algorithms to reduce such data at high throughput to fit permanent storage has drawn increasing attention. Specifically, at the newly constructed sPHENIX experiment at the Relativistic Heavy Ion Collider (RHIC), a time projection chamber is used as the main tracking detector, which records particle trajectories in a volume of three-dimensional (3D) cylinder. The resulting data are usually very sparse with occupancy around 10.8%. Such sparsity presents a challenge to conventional learning-free lossy compression algorithms, such as SZ, ZFP, and MGARD. The 3D convolutional neural network (CNN)-based approach, Bicephalous Convolutional Autoencoder (BCAE), outperforms traditional methods both in compression rate and reconstruction accuracy. BCAE can also utilize the computation power of graphical processing units suitable for deployment in a modern heterogeneous highperformance computing environment. This work introduces two BCAE variants: BCAE++ and BCAE-2D. BCAE++ achieves a 15% better compression ratio and a 77% better reconstruction accuracy measured in mean absolute error compared with BCAE. BCAE-2D treats the radial direction as the channel dimension of an image, resulting in a 3× speedup in compression throughput. In addition, we demonstrate an unbalanced autoencoder with a larger decoder can improve reconstruction accuracy without significantly sacrificing throughput. Lastly, we observe both the BCAE++ and BCAE-2D can benefit more from using half-precision mode in throughput (76 - 79% increase) without loss in reconstruction accuracy. The source code and links to data and pretrained models can be found at https://github.com/BNL-DAQ-LDRD/NeuralCompression_v2

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Cloud and Precipitation Experiment at Kennaook (Cape-K) Science Plan

The Cloud and Precipitation Experiment at Kennaook (CAPE-K) will augment ongoing measurements at the Kennaook/Cape Grim Baseline Air Pollution Station (KCG) with components of the U.S. Department of Energy Atmospheric Radiation Measurement (ARM) user facility’s second Mobile Facility (AMF2) in a deployment that is scheduled to extend from April 2024 through September 2025. Located at Kennaook/Cape Grim on the northwestern tip of Tasmania (40.68° S, 144.69° E), KCG has produced the longest and most consequential record of Southern Hemisphere aerosol and gas-phase chemistry; however, extensive cloud and precipitation measurements have not been collected at this site. CAPE-K is motivated by the fact that model uncertainty in the surface radiation budget due to clouds and precipitation remain significant in this region, where strong latitudinal gradients in cloud feedbacks are found and there are strong and well-documented seasonal variations in marine aerosol properties. Satellite data suggest that cloud properties co-vary with the seasonal cycle in aerosol in the Southern Ocean (SO), yet detailed vertically resolved measurements of cloud and precipitation in the marine boundary layer are sparse. CAPE-K will provide the first such seasonal cycle of measurements in this important latitude band. CAPE-K will provide detailed cloud and precipitation observations over two winter seasons and align with complementary high-value-add activities planned by Australian colleagues, including a voyage of the Research Vessel Investigator (RVI) in May 2025 that will conduct measurements just offshore of KCG.

54 ENVIRONMENTAL SCIENCES↗

UVCGAN: UNet Vision Transformer cycle-consistent GAN for unpaired image-to-image translation

Unpaired image-to-image translation has broad applications in art, design, and scientific simulations. One early breakthrough was CycleGAN that emphasizes one-to-one mappings between two unpaired image domains via generative-adversarial networks (GAN) coupled with the cycle-consistency constraint, while more recent works promote one-to-many mapping to boost diversity of the translated images. Motivated by scientific simulation and one-to-one needs, this work revisits the classic CycleGAN framework and boosts its performance to outperform more contemporary models without relaxing the cycle-consistency constraint. To achieve this, we equip the generator with a Vision Transformer (ViT) and employ necessary training and regularization techniques. Compared to previous best-performing models, our model performs better and retains a strong correlation between the original and translated image. An accompanying ablation study shows that both the gradient penalty and self-supervised pre-training are crucial to the improvement. To promote reproducibility and open science, the source code, hyperparameter configurations, and pre-trained model are available at https: //github.com/LS4GAN/uvcgan.

97 MATHEMATICS AND COMPUTING↗

Snowmass 2021 Computational Frontier CompF03 Topical Group Report: Machine Learning

The rapidly-developing intersection of machine learning (ML) with high-energy physics (HEP) presents both opportunities and challenges to our community. Far beyond applications of standard ML tools to HEP problems, genuinely new and potentially revolutionary approaches are being developed by a generation of talent literate in both fields. There is an urgent need to support the needs of the interdisciplinary community driving these developments, including funding dedicated research at the intersection of the two fields, investing in high-performance computing at universities and tailoring allocation policies to support this work, developing of community tools and standards, and providing education and career paths for young researchers attracted by the intellectual vitality of machine learning for high energy physics.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Probing electronic dead layers in homoepitaxial n-SrTiO 3 (001) films

We combine state-of-the-art oxide epitaxial growth by hybrid molecular beam epitaxy with transport, x-ray photoemission, and surface diffraction, along with classical and first-principles quantum mechanical modeling to investigate the nuances of insulating layer formation in otherwise high-mobility homoepitaxial n-SrTiO 3 (001) films. Our analysis points to charge immobilization at the buried n-SrTiO 3 /undoped SrTiO 3 (001) interface as well as within the surface contamination layer resulting from air exposure as the drivers of electronic dead-layer formation. As Fermi level equilibration occurs at the surface and the buried interface, charge trapping reduces the sheet carrier density (n 2D ) and renders the n-STO film insulating if n 2D falls below the critical value for the metal-to-insulator transition.

36 MATERIALS SCIENCE↗

Scattering mechanisms in state-of-the-art GaAs/AlGaAs quantum wells

Motivated by recent breakthroughs in molecular beam epitaxy of GaAs/AlGaAs quantum wells, we examine contributions to mobility and quantum mobility from various scattering mechanisms and their dependencies on the electron density. Here we find that at lower electron densities, n e ≲ 1 x 10 11 cm -2 , both transport and quantum mobility are limited by unintentional background impurities and follow a power-law dependence, ∝ n$^{α}_{e}$, with α ≈ 0.85. Our predictions for quantum mobility are in reasonable agreement with an estimate obtained from the resistivity at filling factor ν = 1/2 in a sample of Y. J. Chung et al. with n e = 1 x 10 11 cm -2 . Consideration of other scattering mechanisms indicates that interface roughness (remote donors) is likely a limiting factor of transport (quantum) mobility at higher electron densities. Future measurements of quantum mobility should yield information on the distribution of background impurities in GaAs and AlGaAs.

36 MATERIALS SCIENCE↗

Trends in Downwelling Longwave Radiance Over the Southern Great Plains

Downwelling longwave radiation is an important part of the surface energy budget. Spectral trends in the downwelling longwave radiance (DLR) provide insight into the radiative drivers of climate change. In this research, we process and analyze a 23-year DLR record measured by the Atmospheric Emitted Radiance Interferometer (AERI) at the U.S. Department of Energy Atmospheric Radiation Program Southern Great Plains (SGP) site. Two AERIs were deployed at SGP with an overlapping observation period of about 10 years, which allows us to examine the consistency and accuracy of the measurements and to account for discrepancies between them due to errors associated with the instruments themselves. We then analyzed the all-sky radiance trends in DLR, which are associated with the surface warming trend at SGP during this same period and also the complex changes in meteorological conditions. For instance, the observed radiance in the CO2 absorption band follows closely the near-surface air temperature variations. The significant positive radiance trends in weak absorption channels, such as in the wings of the CO 2 band and in the weak absorption channels in the H 2 O vibration-rotational band, show earlier detectability of climate change. The magnitude of the radiance trend uncertainty in the DLR record mainly results from internal climate variability rather than from measurement error, which highlights the importance of continuing the DLR spectral measurements to unambiguously detect and attribute climate change.

54 ENVIRONMENTAL SCIENCES↗

Permafrost thaw with warming reduces microbial metabolic capacities in subsurface soils

Abstract Microorganisms are major constituents of the total biomass in permafrost regions, whose underlain soils are frozen for at least two consecutive years. To understand potential microbial responses to climate change, here we examined microbial community compositions and functional capacities across four soil depths in an Alaska tundra site. We showed that a 5‐year warming treatment increased soil thaw depth by 25.7% ( p = .011) within the deep organic layer (15–25 cm). Concurrently, warming reduced 37% of bacterial abundance and 64% of fungal abundances in the deep organic layer, while it did not affect microbial abundance in other soil layers (i.e., 0–5, 5–15, and 45–55 cm). Warming treatment altered fungal community composition and microbial functional structure ( p < .050), but not bacterial community composition. Using a functional gene array, we found that the relative abundances of a variety of carbon (C)‐decomposing, iron‐reducing, and sulphate‐reducing genes in the deep organic layer were decreased, which was not observed by the shotgun sequencing‐based metagenomics analysis of those samples. To explain the reduced metabolic capacities, we found that warming treatment elicited higher deterministic environmental filtering, which could be linked to water‐saturated time, soil moisture, and soil thaw duration. In contrast, plant factors showed little influence on microbial communities in subsurface soils below 15 cm, despite a 25.2% higher ( p < .05) aboveground plant biomass by warming treatment. Collectively, we demonstrate that microbial metabolic capacities in subsurface soils are reduced, probably arising from enhanced thaw by warming.

Wu, Linwei↗

Efficient Data Compression for 3D Sparse TPC via Bicephalous Convolutional Autoencoder

Real-time data collection and analysis in large experimental facilities present a great challenge across multiple domains, including high energy physics, nuclear physics, and cosmology. To address this, machine learning (ML)-based methods for real-time data compression have drawn significant attention. However, unlike natural image data, such as CIFAR and ImageNet that are relatively small-sized and continuous, scientific data often come in as three-dimensional 3D data volumes at high rates with high sparsity (many zeros) and non-Gaussian value distribution. This makes direct application of popular ML compression methods, as well as conventional data compression methods, suboptimal. To address these obstacles, this work introduces a dual-head autoencoder to resolve sparsity and regression simultaneously, called Bicephalous Convolutional AutoEncoder (BCAE). This method shows advantages both in compression fidelity and ratio compared to traditional data compression methods, such as MGARD, SZ, and ZFP. To achieve similar fidelity, the best performer among the traditional methods can reach only half the compression ratio of BCAE. Moreover, a thorough ablation study of the BCAE method shows that a dedicated segmentation decoder improves the reconstruction.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

An automated multi-component gas adsorption system (MC GAS)

The knowledge gap on adsorption of complex mixtures in the literature relative to single component data represents a persistent obstacle to developing accurate process models for adsorption separations. The collection of mixed gas adsorption data is an imminent need for improved understanding of the behavior of adsorbent systems in these diverse adsorption applications. Current approaches to understanding mixture adsorption using predictive theories based on pure component adsorption experiments often fail to capture the behavior of more complex, non-ideal systems. In this work, we present an automated volumetric instrument for the measurement of mixed gas adsorption isotherms. This instrument was validated by comparison to other in-house instruments and data available in the literature, and the binary adsorption measurements were found to be thermodynamically consistent. The automation of this instrument allows for rapid collection of high-quality mixture adsorption data.

47 OTHER INSTRUMENTATION↗

Observations of Clouds, Aerosols, Precipitation, and Surface Radiation over the Southern Ocean: An Overview of CAPRICORN, MARCUS, MICRE, and SOCRATES

Weather and climate models are challenged by uncertainties and biases in simulating Southern Ocean (SO) radiative fluxes that trace to a poor understanding of cloud, aerosol, precipitation, and radiative processes, and their interactions. Projects between 2016 and 2018 used in situ probes, radar, lidar, and other instruments to make comprehensive measurements of thermodynamics, surface radiation, cloud, precipitation, aerosol, cloud condensation nuclei (CCN), and ice nucleating particles over the SO cold waters, and in ubiquitous liquid and mixed-phase clouds common to this pristine environment. Data including soundings were collected from the NSF–NCAR G-V aircraft flying north–south gradients south of Tasmania, at Macquarie Island, and on the R/V Investigator and RSV Aurora Australis. Synergistically these data characterize boundary layer and free troposphere environmental properties, and represent the most comprehensive data of this type available south of the oceanic polar front, in the cold sector of SO cyclones, and across seasons. Results show largely pristine environments with numerous small and few large aerosols above cloud, suggesting new particle formation and limited long-range transport from continents, high variability in CCN and cloud droplet concentrations, and ubiquitous supercooled water in thin, multilayered clouds, often with small-scale generating cells near cloud top. These observations demonstrate how cloud properties depend on aerosols while highlighting the importance of dynamics and turbulence that likely drive heterogeneity of cloud phase. Satellite retrievals confirmed low clouds were responsible for radiation biases. Furthermore, the combination of models and observations is examining how aerosols and meteorology couple to control SO water and energy budgets.

54 ENVIRONMENTAL SCIENCES↗

Microbial Functional Responses Explain Alpine Soil Carbon Fluxes under Future Climate Scenarios

Soil microorganisms are sensitive to temperature in cold ecosystems, but it remains unclear how microbial responses are modulated by other important climate drivers, such as precipitation changes. Here, we examine the effects of six in situ warming and/or precipitation treatments in alpine grasslands on microbial communities, plants, and soil carbon fluxes. These treatments differentially affected soil carbon fluxes, gross primary production, and microbial communities. Variations of soil CO 2 and CH 4 fluxes across all sites significantly (r > 0.70, P < 0.050) correlated with relevant microbial functional abundances but not bacterial or fungal abundances. Given tight linkages between microbial functional traits and ecosystem functionality, we conclude that future soil carbon fluxes in alpine grasslands can be predicted by microbial carbon-degrading capacities.

54 ENVIRONMENTAL SCIENCES↗