arXiv subjects

Daniel Murnane

Publications and source records attributed to Daniel Murnane.

At least 19 recordsLinked to original sources

arXiv2026

chATLAS: A Retrieval-Augmented Conversational Assistant for the ATLAS Experiment

Large scientific collaborations produce extensive internal documentation spread across heterogeneous systems, making it difficult for collaborators to locate and reuse existing knowledge. We present chATLAS, a deployed Retrieval-Augmented Generation (RAG) assistant developed for the ATLAS experiment at CERN. The system provides conversational access to more than 30,000 documents from several collaboration resources, material largely unavailable to general-purpose language models, using hybrid dense--sparse retrieval deployed on CERN infrastructure. A domain-specific benchmark built from automatically generated question--answer pairs is used to evaluate retrieval and generation performance and to guide design choices. We study the selection and domain-specific fine-tuning of an embedding model, contextual retrieval, ranking strategies, and generator choice. Fine-tuning combined with contextual retrieval more than doubles how often the correct document is returned as the top result on the most technically challenging documentation source. The deployed system also includes intent routing, safety handling, Mattermost integration, and an MCP server for future agentic workflows. chATLAS demonstrates that domain-adapted RAG can provide source-attributed, domain-specific access to collaboration knowledge, and offers a practical foundation for scientific collaborations and other organisations that rely on accurate, up-to-date retrieval from proprietary documentation.

cs.IR↗

arXiv2026

Precision Tokenization for Multi-Modal Data in High Energy Physics

Foundation models have recently gained significant attention in fields such as high energy physics (HEP). The tokenizers that convert detector data into discrete inputs for such models, however, are comparatively unexplored. This paper addresses this gap by learning discrete and compressed tokenized representations of HEP datasets. We demonstrate the approach on four modalities covering two different simulated detectors that vary in complexity. Treating tokenization as a rate-distortion problem, a systematic hyperparameter scan yields four design rules for tokenizing HEP data that hold across all set-based modalities studied. The cost of tokenization is evaluated on tasks including jet $p_T$ regression, particle flow reconstruction, and flavor tagging. We find that tokenization preserves much of the information needed for these tasks at compression factors of approximately 4-11$\times$ while keeping the jet $p_T$ resolution below 1%. We also present heptokens, an open-source library for tokenizing HEP data.

hep-ph↗

arXiv2026

Fast and Precise Learned Charged-Particle Trajectory Regression at the Large Hadron Collider

We propose a training recipe that treats charged-particle trajectory parameter regression on high-energy physics detector data as a sequence-modeling task. Kalman filters and linearized least-squares fits have been the classical standard approach for this task: they are optimal estimators for sparsely sampled linear-Gaussian data and are commonly used for trajectory parameter regression (fitting). The classical fitting techniques implemented for this domain reach a final precision of one part in $10^5$ through detailed modeling of detector geometry, material, detection effects and precise numerical integration of the equations of motion through the detector's inhomogeneous magnetic field. With this study, we demonstrate that using a bidirectional gated linear recurrent encoder, one is able to reproduce the full precision of classical track fitting techniques. Using a custom kernel, we also achieve significantly higher throughput during GPU inference, compared to classical fitting software running on similarly priced multi-core CPU servers representing typically employed hardware. Such a speedup would lead to considerable cost savings for the pattern recognition at the Large Hadron Collider. To our knowledge, this is the first end-to-end learned track fit to reach the full precision and, at the same time, offer the opportunity to reduce the computing costs.

cs.LG↗

arXiv2026

Prompting Particle Physics: Tokenized Multi-modal Foundation Models for Combinatorially Many Tasks

Reconstruction and simulation at a collider experiment are long chains of specialised algorithms, each tuned to a single step. We explore how one model can serve many of those steps at once, while still producing the intermediate objects (tracks, calorimeter cells, clusters, particles and jets) that make the chain interpretable. To do so, we represent every object in a jet in a single shared token vocabulary and train one model to map any subset of these modalities to any other. A task is then only a choice of which modalities to provide and which to request: particle flow, detector simulation and charged energy subtraction are all directions through the same set of weights. We train over all modality combinations, with a decoder emitting tokens either autoregressively or in parallel. With tokenisation, both architectures train stably with little tuning. In evaluations, both models produce realistic reconstruction and simulation objects, with the autoregressive model particularly faithful to output from Geant4. The autoregressive model also outperforms a state-of-the-art particle flow algorithm on many typical jet reconstruction metrics.

hep-ph↗

arXiv2026

The Landscape of Composite Higgs Models

While the Standard Model (SM) of particle physics contains the most precise set of predictions ever devised by humanity, that precision comes at a cost. The strange nature of the Higgs particle requires its parameters to be tuned so precisely that if the SM is indeed the true description of reality, one is forced to wonder how such a miracle as galactic structure and life could occur. Instead, we search in this work for a natural explanation. The concept of naturalness is comprehensively explored, and a new tuning measure proposed, with an aim to place it on well-defined Bayesian footing. We then turn this measure on to the analysis of a class of intriguing new physics - Composite Higgs models. These effective models are the result of a plethora of underlying theories, and they allow the production of a naturally light Higgs particle, appearing as the SM Higgs at low energy. We establish the background required to appreciate the N-site 4D Composite Higgs model, and subsequently focus on the simplest incarnations of this class. A global fit is performed on the Minimal 4D Composite Higgs model (M4DCHM), with strong exclusion bounds placed on collider search channels. We analyse any improvement in tuning that could be gained from several extensions to this model. The Leptonic M4DCHM is explored, with a composite tau lepton embedded in various representations. The possibility of a dark matter candidate existing in the Next-to-Minimal 4DCHM is considered. Ultimately, we are able to define what, if any, benefit to naturalness can come to the Composite Higgs sector by introducing these extensions.

hep-ph↗

arXiv2025

ColliderML: The First Release of an OpenDataDetector High-Luminosity Physics Benchmark Dataset

We introduce ColliderML - a large, open, experiment-agnostic dataset of fully simulated and digitised proton-proton collisions in High-Luminosity Large Hadron Collider conditions ($\sqrt{s}=14$ TeV, mean pile-up $μ= 200$). ColliderML provides one million events across ten Standard Model and Beyond Standard Model processes, plus extensive single-particle samples, all produced with modern next-to-leading order matrix element calculation and showering, realistic per-event pile-up overlay, a validated OpenDataDetector geometry, and standard reconstructions. The release fills a major gap for machine learning (ML) research on detector-level data, provided on the ML-friendly Hugging Face platform. We present physics coverage and the generation, simulation, digitisation and reconstruction pipeline, describe format and access, and initial collider physics benchmarks.

hep-ex↗

arXiv2025

Learning to Reconstruct Quirky Tracks

Analysis of data from particle physics experiments traditionally sacrifices some sensitivity to new particles for the sake of practical computability, effectively ignoring some potentially striking signatures. However, recent advances in ML-based tracking allow for new inroads into previously inaccessible territory, such as reconstruction of tracks which do not follow helical trajectories. This paper presents a demonstration of the capacity of ML-based tracking to reconstruct the oscillating trajectories of quirks. The technique used is not specific to quirks, and opens the door to a program of searching for many kinds of non-standard tracks.

hep-ex↗

arXiv2025

Machine Learning Power Week 2023: Clustering in Hadronic Calorimeters

In both high-energy physics and industry applications, a crowd-sourced approach to difficult problems is becoming increasingly common. These innovative approaches are ideal for the development of future facilities where the simulations can be publicly distributed, such as the Electron-Ion Collider (EIC). In this paper, we discuss a so-called ``Power Week" where graduate students were able to learn about machine learning while also contributing to an unsolved problem at a future facility. Here, the problem of interest was the clustering of the forward hadronic calorimeter in the foreseen electron-proton/ion collider experiment (ePIC) detector at the EIC. The different possible approaches, developed over the course of a single week, and their performance are detailed and summarised. Feedback on the format of the week and recommendations for future similar programs are provided in the hopes to inspire future learning opportunities for students that also serve as a crowd-sourced approaches to unsolved problems.

nucl-ex↗

arXiv2023

Equivariance Is Not All You Need: Characterizing the Utility of Equivariant Graph Neural Networks for Particle Physics Tasks

Incorporating inductive biases into ML models is an active area of ML research, especially when ML models are applied to data about the physical world. Equivariant Graph Neural Networks (GNNs) have recently become a popular method for learning from physics data because they directly incorporate the symmetries of the underlying physical system. Drawing from the relevant literature around group equivariant networks, this paper presents a comprehensive evaluation of the proposed benefits of equivariant GNNs by using real-world particle physics reconstruction tasks as an evaluation test-bed. We demonstrate that many of the theoretical benefits generally associated with equivariant networks may not hold for realistic systems and introduce compelling directions for future research that will benefit both the scientific theory of ML and physics applications.

cs.LG↗

arXiv2023

Graph Structure from Point Clouds: Geometric Attention is All You Need

The use of graph neural networks has produced significant advances in point cloud problems, such as those found in high energy physics. The question of how to produce a graph structure in these problems is usually treated as a matter of heuristics, employing fully connected graphs or K-nearest neighbors. In this work, we elevate this question to utmost importance as the Topology Problem. We propose an attention mechanism that allows a graph to be constructed in a learned space that handles geometrically the flow of relevance, providing one solution to the Topology Problem. We test this architecture, called GravNetNorm, on the task of top jet tagging, and show that it is competitive in tagging accuracy, and uses far fewer computational resources than all other comparable models.

cs.LG↗

arXiv2023

Heterogeneous Graph Neural Network for Identifying Hadronically Decayed Tau Leptons at the High Luminosity LHC

We present a new algorithm that identifies reconstructed jets originating from hadronic decays of tau leptons against those from quarks or gluons. No tau lepton reconstruction algorithm is used. Instead, the algorithm represents jets as heterogeneous graphs with tracks and energy clusters as nodes and trains a Graph Neural Network to identify tau jets from other jets. Different attributed graph representations and different GNN architectures are explored. We propose to use differential track and energy cluster information as node features and a heterogeneous sequentially-biased encoding for the inputs to final graph-level classification.

physics.ins-det↗

arXiv2023

Equivariant Graph Neural Networks for Charged Particle Tracking

Graph neural networks (GNNs) have gained traction in high-energy physics (HEP) for their potential to improve accuracy and scalability. However, their resource-intensive nature and complex operations have motivated the development of symmetry-equivariant architectures. In this work, we introduce EuclidNet, a novel symmetry-equivariant GNN for charged particle tracking. EuclidNet leverages the graph representation of collision events and enforces rotational symmetry with respect to the detector's beamline axis, leading to a more efficient model. We benchmark EuclidNet against the state-of-the-art Interaction Network on the TrackML dataset, which simulates high-pileup conditions expected at the High-Luminosity Large Hadron Collider (HL-LHC). Our results show that EuclidNet achieves near-state-of-the-art performance at small model scales (<1000 parameters), outperforming the non-equivariant benchmarks. This study paves the way for future investigations into more resource-efficient GNN models for particle tracking in HEP experiments.

physics.ins-det↗

arXiv2022

Benchmarking GPU and TPU Performance with Graph Neural Networks

Many artificial intelligence (AI) devices have been developed to accelerate the training and inference of neural networks models. The most common ones are the Graphics Processing Unit (GPU) and Tensor Processing Unit (TPU). They are highly optimized for dense data representations. However, sparse representations such as graphs are prevalent in many domains, including science. It is therefore important to characterize the performance of available AI accelerators on sparse data. This work analyzes and compares the GPU and TPU performance training a Graph Neural Network (GNN) developed to solve a real-life pattern recognition problem. Characterizing the new class of models acting on sparse data may prove helpful in optimizing the design of deep learning libraries and future AI accelerators.

cs.LG↗

arXiv2022

Graph Neural Networks in Particle Physics: Implementations, Innovations, and Challenges

Many physical systems can be best understood as sets of discrete data with associated relationships. Where previously these sets of data have been formulated as series or image data to match the available machine learning architectures, with the advent of graph neural networks (GNNs), these systems can be learned natively as graphs. This allows a wide variety of high- and low-level physical features to be attached to measurements and, by the same token, a wide variety of HEP tasks to be accomplished by the same GNN architectures. GNNs have found powerful use-cases in reconstruction, tagging, generation and end-to-end analysis. With the wide-spread adoption of GNNs in industry, the HEP community is well-placed to benefit from rapid improvements in GNN latency and memory usage. However, industry use-cases are not perfectly aligned with HEP and much work needs to be done to best match unique GNN capabilities to unique HEP obstacles. We present here a range of these capabilities, predictions of which are currently being well-adopted in HEP communities, and which are still immature. We hope to capture the landscape of graph techniques in machine learning as well as point out the most significant gaps that are inhibiting potentially large leaps in research.

hep-ex↗

arXiv2022

Reconstruction of Large Radius Tracks with the Exa.TrkX pipeline

Particle tracking is a challenging pattern recognition task at the Large Hadron Collider (LHC) and the High Luminosity-LHC. Conventional algorithms, such as those based on the Kalman Filter, achieve excellent performance in reconstructing the prompt tracks from the collision points. However, they require dedicated configuration and additional computing time to efficiently reconstruct the large radius tracks created away from the collision points. We developed an end-to-end machine learning-based track finding algorithm for the HL-LHC, the Exa.TrkX pipeline. The pipeline is designed so as to be agnostic about global track positions. In this work, we study the performance of the Exa.TrkX pipeline for finding large radius tracks. Trained with all tracks in the event, the pipeline simultaneously reconstructs prompt tracks and large radius tracks with high efficiencies. This new capability offered by the Exa.TrkX pipeline may enable us to search for new physics in real time.

physics.ins-det↗

arXiv2022

Symmetry Group Equivariant Architectures for Physics

Physical theories grounded in mathematical symmetries are an essential component of our understanding of a wide range of properties of the universe. Similarly, in the domain of machine learning, an awareness of symmetries such as rotation or permutation invariance has driven impressive performance breakthroughs in computer vision, natural language processing, and other important applications. In this report, we argue that both the physics community and the broader machine learning community have much to understand and potentially to gain from a deeper investment in research concerning symmetry group equivariant machine learning architectures. For some applications, the introduction of symmetries into the fundamental structural design can yield models that are more economical (i.e. contain fewer, but more expressive, learned parameters), interpretable (i.e. more explainable or directly mappable to physical quantities), and/or trainable (i.e. more efficient in both data and computational requirements). We discuss various figures of merit for evaluating these models as well as some potential benefits and limitations of these methods for a variety of physics applications. Research and investment into these approaches will lay the foundation for future architectures that are potentially more robust under new computational paradigms and will provide a richer description of the physical systems to which they are applied.

cs.LG↗

arXiv2022

Accelerating the Inference of the Exa.TrkX Pipeline

Recently, graph neural networks (GNNs) have been successfully used for a variety of particle reconstruction problems in high energy physics, including particle tracking. The Exa.TrkX pipeline based on GNNs demonstrated promising performance in reconstructing particle tracks in dense environments. It includes five discrete steps: data encoding, graph building, edge filtering, GNN, and track labeling. All steps were written in Python and run on both GPUs and CPUs. In this work, we accelerate the Python implementation of the pipeline through customized and commercial GPU-enabled software libraries, and develop a C++ implementation for inferencing the pipeline. The implementation features an improved, CUDA-enabled fixed-radius nearest neighbor search for graph building and a weakly connected component graph algorithm for track labeling. GNNs and other trained deep learning models are converted to ONNX and inferenced via the ONNX Runtime C++ API. The complete C++ implementation of the pipeline allows integration with existing tracking software. We report the memory usage and average event latency tracking performance of our implementation applied to the TrackML benchmark dataset.

physics.ins-det↗

arXiv2022

Semi-Equivariant GNN Architectures for Jet Tagging

Composing Graph Neural Networks (GNNs) of operations that respect physical symmetries has been suggested to give better model performance with a smaller number of learnable parameters. However, real-world applications, such as in high energy physics have not born this out. We present the novel architecture VecNet that combines both symmetry-respecting and unconstrained operations to study and tune the degree of physics-informed GNNs. We introduce a novel metric, the \textit{ant factor}, to quantify the resource-efficiency of each configuration in the search-space. We find that a generalized architecture such as ours can deliver optimal performance in resource-constrained applications.

hep-ph↗