arXiv subjects

Elham E Khoda

Publications and source records attributed to Elham E Khoda.

17 recordsLinked to original sources

arXiv2026

Spatially Aware Linear Transformer (SAL-T) for Particle Jet Tagging

Transformers are very effective in capturing both global and local correlations within high-energy particle collisions, but they present deployment challenges in high-data-throughput environments, such as the CERN LHC. The quadratic complexity of transformer models demands substantial resources and increases latency during inference. In order to address these issues, we introduce the Spatially Aware Linear Transformer (SAL-T), a physics-inspired enhancement of the linformer architecture that maintains linear attention. Our method incorporates spatially aware partitioning of particles based on kinematic features, thereby computing attention between regions of physical significance. Additionally, we employ convolutional layers to capture local correlations, informed by insights from jet physics. In addition to outperforming the standard linformer in jet classification tasks, SAL-T also achieves classification results comparable to full-attention transformers, while using considerably fewer resources with lower latency during inference. Experiments on a generic point cloud classification dataset (ModelNet10) further confirm this trend. Our code is available at https://github.com/aaronw5/SAL-T4HEP.

cs.LG↗

arXiv2026

FAIR Universe Weak Lensing ML Uncertainty Challenge: Handling Uncertainties and Distribution Shifts for Precision Cosmology

Weak gravitational lensing, the correlated distortion of background galaxy shapes by foreground structures, is a powerful probe of the matter distribution in our universe and allows accurate constraints on the cosmological model. In recent years, high-order statistics and machine learning (ML) techniques have been applied to weak lensing data to extract the nonlinear information beyond traditional two-point analysis. However, these methods typically rely on cosmological simulations, which poses several challenges: simulations are computationally expensive, limiting most realistic setups to a low training data regime; inaccurate modeling of systematics in the simulations create distribution shifts that can bias cosmological parameter constraints; and varying simulation setups across studies make method comparison difficult. To address these difficulties, we present the first weak lensing benchmark dataset with several realistic systematics and launch the FAIR Universe Weak Lensing Machine Learning Uncertainty Challenge. The challenge focuses on measuring the fundamental properties of the universe from weak lensing data with limited training set and potential distribution shifts, while providing a standardized benchmark for rigorous comparison across methods. Organized in two phases, the challenge will bring together the physics and ML communities to advance the methodologies for handling systematic uncertainties, data efficiency, and distribution shifts in weak lensing analysis with ML, ultimately facilitating the deployment of ML approaches into upcoming weak lensing survey analysis.

astro-ph.CO↗

arXiv2025

wa-hls4ml: A Benchmark and Surrogate Models for hls4ml Resource and Latency Estimation

As machine learning (ML) is increasingly implemented in hardware to address real-time challenges in scientific applications, the development of advanced toolchains has significantly reduced the time required to iterate on various designs. These advancements have solved major obstacles, but also exposed new challenges. For example, processes that were not previously considered bottlenecks, such as hardware synthesis, are becoming limiting factors in the rapid iteration of designs. To mitigate these emerging constraints, multiple efforts have been undertaken to develop an ML-based surrogate model that estimates resource usage of ML accelerator architectures. We introduce wa-hls4ml, a benchmark for ML accelerator resource and latency estimation, and its corresponding initial dataset of over 680,000 fully connected and convolutional neural networks, all synthesized using hls4ml and targeting Xilinx FPGAs. The benchmark evaluates the performance of resource and latency predictors against several common ML model architectures, primarily originating from scientific domains, as exemplar models, and the average performance across a subset of the dataset. Additionally, we introduce GNN- and transformer-based surrogate models that predict latency and resources for ML accelerators. We present the architecture and performance of the models and find that the models generally predict latency and resources for the 75% percentile within several percent of the synthesized resources on the synthetic test dataset.

cs.LG↗

arXiv2025

Serving LLMs in HPC Clusters: A Comparative Study of Qualcomm Cloud AI 100 Ultra and NVIDIA Data Center GPUs

This study presents a benchmarking analysis of the Qualcomm Cloud AI 100 Ultra (QAic) accelerator for large language model (LLM) inference, evaluating its energy efficiency (throughput per watt), performance, and hardware scalability against NVIDIA A100 GPUs (in 4x and 8x configurations) within the National Research Platform (NRP) ecosystem. A total of 12 open-source LLMs, ranging from 124 million to 70 billion parameters, are served using the vLLM framework. Our analysis reveals that QAic achieves competitive energy efficiency with advantages on specific models while enabling more granular hardware allocation: some 70B models operate on as few as 1 QAic card versus 8 A100 GPUs required, with 20x lower power consumption (148W vs 2,983W). For smaller models, single QAic devices achieve up to 35x lower power consumption compared to our 4-GPU A100 configuration (36W vs 1,246W). The findings offer insights into the potential of the Qualcomm Cloud AI 100 Ultra for energy-constrained and resource-efficient HPC deployments within the National Research Platform (NRP).

cs.DC↗

arXiv2025

The National Research Platform: Stretched, Multi-Tenant, Scientific Kubernetes Cluster

The National Research Platform (NRP) represents a distributed, multi-tenant Kubernetes-based cyberinfrastructure designed to facilitate collaborative scientific computing. Spanning over 75 locations in the U.S. and internationally, the NRP uniquely integrates varied computational resources, ranging from single nodes to extensive GPU and CPU clusters, to support diverse research workloads including advanced AI and machine learning tasks. It emphasizes flexibility through user-friendly interfaces such as JupyterHub and low level control of resources through direct Kubernetes interaction. Critical operational insights are discussed, including security enhancements using Kubernetes-integrated threat detection, extensive monitoring, and comprehensive accounting systems. This paper highlights the NRP's growing importance and scalability in addressing the increasing demands for distributed scientific computational resources.

cs.DC↗

arXiv2024

Interpreting Transformers for Jet Tagging

Machine learning (ML) algorithms, particularly attention-based transformer models, have become indispensable for analyzing the vast data generated by particle physics experiments like ATLAS and CMS at the CERN LHC. Particle Transformer (ParT), a state-of-the-art model, leverages particle-level attention to improve jet-tagging tasks, which are critical for identifying particles resulting from proton collisions. This study focuses on interpreting ParT by analyzing attention heat maps and particle-pair correlations on the $η$-$ϕ$ plane, revealing a binary attention pattern where each particle attends to at most one other particle. At the same time, we observe that ParT shows varying focus on important particles and subjets depending on decay, indicating that the model learns traditional jet substructure observables. These insights enhance our understanding of the model's internal workings and learning process, offering potential avenues for improving the efficiency of transformer architectures in future high-energy physics applications.

hep-ph↗

arXiv2024

Low Latency Transformer Inference on FPGAs for Physics Applications with hls4ml

This study presents an efficient implementation of transformer architectures in Field-Programmable Gate Arrays(FPGAs) using hls4ml. We demonstrate the strategy for implementing the multi-head attention, softmax, and normalization layer and evaluate three distinct models. Their deployment on VU13P FPGA chip achieved latency less than 2us, demonstrating the potential for real-time applications. HLS4ML compatibility with any TensorFlow-built transformer model further enhances the scalability and applicability of this work. Index Terms: FPGAs, machine learning, transformers, high energy physics, LIGO

cs.LG↗

arXiv2024

Graph Neural Network-based Tracking as a Service

Recent studies have shown promising results for track finding in dense environments using Graph Neural Network (GNN)-based algorithms. However, GNN-based track finding is computationally slow on CPUs, necessitating the use of coprocessors to accelerate the inference time. Additionally, the large input graph size demands a large device memory for efficient computation, a requirement not met by all computing facilities used for particle physics experiments, particularly those lacking advanced GPUs. Furthermore, deploying the GNN-based track-finding algorithm in a production environment requires the installation of all dependent software packages, exclusively utilized by this algorithm. These computing challenges must be addressed for the successful implementation of GNN-based track-finding algorithm into production settings. In response, we introduce a ``GNN-based tracking as a service'' approach, incorporating a custom backend within the NVIDIA Triton inference server to facilitate GNN-based tracking. This paper presents the performance of this approach using the Perlmutter supercomputer at NERSC.

physics.comp-ph↗

arXiv2024

FPGA Deployment of LFADS for Real-time Neuroscience Experiments

Large-scale recordings of neural activity are providing new opportunities to study neural population dynamics. A powerful method for analyzing such high-dimensional measurements is to deploy an algorithm to learn the low-dimensional latent dynamics. LFADS (Latent Factor Analysis via Dynamical Systems) is a deep learning method for inferring latent dynamics from high-dimensional neural spiking data recorded simultaneously in single trials. This method has shown a remarkable performance in modeling complex brain signals with an average inference latency in milliseconds. As our capacity of simultaneously recording many neurons is increasing exponentially, it is becoming crucial to build capacity for deploying low-latency inference of the computing algorithms. To improve the real-time processing ability of LFADS, we introduce an efficient implementation of the LFADS models onto Field Programmable Gate Arrays (FPGA). Our implementation shows an inference latency of 41.97 $μ$s for processing the data in a single trial on a Xilinx U55C.

q-bio.NC↗

arXiv2024

Ultra Fast Transformers on FPGAs for Particle Physics Experiments

This work introduces a highly efficient implementation of the transformer architecture on a Field-Programmable Gate Array (FPGA) by using the \texttt{hls4ml} tool. Given the demonstrated effectiveness of transformer models in addressing a wide range of problems, their application in experimental triggers within particle physics becomes a subject of significant interest. In this work, we have implemented critical components of a transformer model, such as multi-head attention and softmax layers. To evaluate the effectiveness of our implementation, we have focused on a particle physics jet flavor tagging problem, employing a public dataset. We recorded latency under 2 $μ$s on the Xilinx UltraScale+ FPGA, which is compatible with hardware trigger requirements at the CERN Large Hadron Collider experiments.

cs.LG↗

arXiv2023

Anomalous production of massive gauge boson pairs at muon colliders

The prospects of searches for anomalous production of hadronically decaying weak boson pairs at proposed high-energy muon colliders are reported. Muon-muon collision events are simulated at $\sqrt{s}=6$, 10, and 30 TeV, corresponding to an integrated luminosity of $4$, $10$, and $10$ ab$^{-1}$, respectively. Simulated $μμ\rightarrow\mathrm{W}\mathrm{W}+νν/μμ$ events are used to set expected constraints on the structure of quartic vector boson interactions in the framework of a dimension-8 effective field theory. Similarly, $μμ\rightarrow\mathrm{W}\mathrm{W}/\mathrm{Z}\mathrm{Z}+νν$ events are used to report constraints on the product of the cross section and branching fraction for vector boson fusion production of a heavy neutral Higgs boson decaying to weak boson pairs. These results are interpreted in the context of the Georgi-Machacek model.

hep-ex↗

arXiv2023

Applications of Deep Learning to physics workflows

Modern large-scale physics experiments create datasets with sizes and streaming rates that can exceed those from industry leaders such as Google Cloud and Netflix. Fully processing these datasets requires both sufficient compute power and efficient workflows. Recent advances in Machine Learning (ML) and Artificial Intelligence (AI) can either improve or replace existing domain-specific algorithms to increase workflow efficiency. Not only can these algorithms improve the physics performance of current algorithms, but they can often be executed more quickly, especially when run on coprocessors such as GPUs or FPGAs. In the winter of 2023, MIT hosted the Accelerating Physics with ML at MIT workshop, which brought together researchers from gravitational-wave physics, multi-messenger astrophysics, and particle physics to discuss and share current efforts to integrate ML tools into their workflows. The following white paper highlights examples of algorithms and computing frameworks discussed during this workshop and summarizes the expected computing needs for the immediate future of the involved fields.

hep-ex↗

arXiv2022

Report of the Topical Group on Electroweak Precision Physics and Constraining New Physics for Snowmass 2021

The precise measurement of physics observables and the test of their consistency within the standard model (SM) are an invaluable approach, complemented by direct searches for new particles, to determine the existence of physics beyond the standard model (BSM). Studies of massive electroweak gauge bosons (W and Z bosons) are a promising target for indirect BSM searches, since the interactions of photons and gluons are strongly constrained by the unbroken gauge symmetries. They can be divided into two categories: (a) Fermion scattering processes mediated by s- or t-channel W/Z bosons, also known as electroweak precision measurements; and (b) multi-boson processes, which include production of two or more vector bosons in fermion-antifermion annihilation, as well as vector boson scattering (VBS) processes. The latter categories can test modifications of gauge-boson self-interactions, and the sensitivity is typically improved with increased collision energy. This report evaluates the achievable precision of a range of future experiments, which depend on the statistics of the collected data sample, the experimental and theoretical systematic uncertainties, and their correlations. In addition it presents a combined interpretation of these results, together with similar studies in the Higgs and top sector, in the Standard Model effective field theory (SMEFT) framework. This framework provides a model-independent prescription to put generic constraints on new physics and to study and combine large sets of experimental observables, assuming that the new physics scales are significantly higher than the EW scale.

hep-ph↗

arXiv2022

Ultra-low latency recurrent neural network inference on FPGAs for physics applications with hls4ml

Recurrent neural networks have been shown to be effective architectures for many tasks in high energy physics, and thus have been widely adopted. Their use in low-latency environments has, however, been limited as a result of the difficulties of implementing recurrent architectures on field-programmable gate arrays (FPGAs). In this paper we present an implementation of two types of recurrent neural network layers -- long short-term memory and gated recurrent unit -- within the hls4ml framework. We demonstrate that our implementation is capable of producing effective designs for both small and large models, and can be customized to meet specific design requirements for inference latencies and FPGA resources. We show the performance and synthesized designs for multiple neural networks, many of which are trained specifically for jet identification tasks at the CERN Large Hadron Collider.

cs.LG↗

arXiv2022

Physics Community Needs, Tools, and Resources for Machine Learning

Machine learning (ML) is becoming an increasingly important component of cutting-edge physics research, but its computational requirements present significant challenges. In this white paper, we discuss the needs of the physics community regarding ML across latency and throughput regimes, the tools and resources that offer the possibility of addressing these needs, and how these can be best utilized and accessed in the coming years.

cs.LG↗

arXiv2022

Study of Electroweak Phase Transition in Exotic Higgs Decays at the CEPC

A strong first-order electroweak phase transition (EWPT) can be induced by light new physics weakly coupled to the Higgs. This study focuses on a scenario in which the first-order EWPT is driven by a light scalar $s$ with a mass between 15-60 GeV. A search for exotic decays of the Higgs boson into a pair of spin-zero particles, $h \to ss$, where the $s$-boson decays into $b$-quarks promptly is presented. The search is performed in events where the Higgs boson is produced in association with a $Z$ boson, giving rise to a signature of two charged leptons (electrons or muons) and multiple jets from $b$-quark decays. The analysis is considering a scenario of analysing 5000 fb$^{-1}$ $e^+ e^-$ collision data at $\sqrt{s} = 240 $ GeV from the Circular Electron Positron Collider (CEPC). This study with $4b$ final state conclusively tests the expected sensitivity of probing the light scalars in the CEPC experiment. The sensitivity reach is significantly larger than that can be achieved at the LHC.

hep-ex↗

arXiv2021

Applications and Techniques for Fast Machine Learning in Science

In this community review report, we discuss applications and techniques for fast machine learning (ML) in science -- the concept of integrating power ML methods into the real-time experimental data processing loop to accelerate scientific discovery. The material for the report builds on two workshops held by the Fast ML for Science community and covers three main areas: applications for fast ML across a number of scientific domains; techniques for training and implementing performant and resource-efficient ML algorithms; and computing architectures, platforms, and technologies for deploying these algorithms. We also present overlapping challenges across the multiple scientific domains where common solutions can be found. This community report is intended to give plenty of examples and inspiration for scientific discovery through integrated and accelerated ML solutions. This is followed by a high-level overview and organization of technical advances, including an abundance of pointers to source material, which can enable these breakthroughs.

cs.LG↗