Mohammad Atif, Kriti Chopra, Fang-Ying Tsai
et al.
Large Language Models (LLM) are increasingly used for software development, yet existing benchmarks for LLM-based coding assistance do not reflect the constraints of High Energy Physics (HEP) and High Performance Computing (HPC) software. Code correctness must respect science constraints and changes must integrate into large, performance-critical codebases with complex dependencies and build systems. The primary contribution of this paper is the development of practical, repeatable benchmarks that quantify LLM performance on HEP/HPC-relevant tasks. We introduce three evaluation tracks -- code documentation benchmarks measure the ability of an LLM to generate Doxygen-style comments, code generation benchmarks evaluate end-to-end usability on representative GPU kernels, and graphical data analysis benchmarks evaluate vision-enabled LLMs. These benchmarks provide a unified framework for measuring progress in scientific coding assistance across documentation quality, code generation robustness, and multimodal validation analysis. By emphasizing repeatability, automated scoring, and domain-relevant failure modes, the suite enables fair comparisons of models and settings while supporting future work on methods that improve reliability for HEP/HPC software development.
Syed Haider Ali, Ashfaq Ahmad, Muhammad Saiel
et al.
The simultaneous production of four top quarks in association with a ($W$) boson at $(\sqrt{s} = 13)$ TeV is an rare SM process with a next-to-leading-order (NLO) cross-section of $(6.6^{+2.4}_{-2.6} {ab})$\cite{saiel}. Identifying this process in the fully hadronic decay channel is particularly challenging due to overwhelming backgrounds from $t\bar{t}, t\bar{t}W, t\bar{t}Z$, and triple-top production processes. This study introduces a modified physics informed Neural Network, a hybrid graph neural network (GNN) enhancing event classification. The proposed model integrates Graph layers for particle-level features, a custom Multi Layer Perceptron(MLP) based global stream with a quantum circuit and cross-attention fusion to combine local and global representations. Physics-informed Loss function enforce jet multiplicity constraints, derived from event decay dynamics. Benchmarked against conventional methods, the GNN achieves a signal significance $(S/\sqrt{S+B})$ of $0.174$ and ROC-AUC of 0.974, surpassing BDT's significance of $0.148$ and ROC of $0.913$, while Xgboost achieves a significance of $0.149$ and ROC of $0.920$. The classification models are trained on Monte Carlo (MC) simulations, with events normalized using cross-section-based reweighting to reflect their expected contributions in a dataset corresponding to $350\;$fb$^{-1}$ of integrated luminosity. This enhanced approach offers a framework for precision event selection at the LHC, leveraging high dimensional statistical learning and physics informed inference to tackle fundamental HEP challenges, aligning with ML developments.
Simone Campana, Irakli Chakaberia, Gang Chen
et al.
These recommendations are the result of reflections by scientists and experts who are, or have been, involved in the preservation of high-energy physics data. The work has been done under the umbrella of the Data Lifecycle panel of the International Committee of Future Accelerators (ICFA), drawing on the expertise of a wide range of stakeholders. A key indicator of success in the data preservation efforts is the long-term usability of the data. Experience shows that achieving this requires providing a rich set of information in various forms, which can only be effectively collected and preserved during the period of active data use. The recommendations are intended to be actionable by the indicated actors and specific to the particle physics domain. They cover a wide range of actions, many of which are interdependent. These dependencies are indicated within the recommendations and can be used as a road map to guide implementation efforts. These recommendations are best accessed and viewed through the web application, see https://icfa-data-best-practices.app.cern.ch/
The second PyHEP.dev workshop, part of the "Python in HEP Developers" series organized by the HEP Software Foundation (HSF), took place in Aachen, Germany, from August 26 to 30, 2024. This gathering brought together nearly 30 Python package developers, maintainers, and power users to engage in informal discussions about current trends in Python, with a primary focus on analysis tools and techniques in High Energy Physics (HEP). The workshop agenda encompassed a range of topics, such as defining the scope of HEP data analysis, exploring the Analysis Grand Challenge project, evaluating statistical models and serialization methods, assessing workflow management systems, examining histogramming practices, and investigating distributed processing tools like RDataFrame, Coffea, and Dask. Additionally, the workshop dedicated time to brainstorming the organization of future PyHEP.dev events, upholding the tradition of alternating between Europe and the United States as host locations. This document, prepared by the session conveners in the weeks following the workshop, serves as a summary of the key discussions, salient points, and conclusions that emerged.
P. B. Cushman, M. C. Fritts, A. D. Chambers
et al.
Background reduction in the SuperCDMS dark matter experiment depends on removing surface events within individual detectors by identifying the location of each incident particle interaction. Position reconstruction is achieved by combining pulse shape information over multiple phonon channels, a task well-suited to machine learning techniques. Data from an Am-241 scan of a SuperCDMS SNOLAB detector was used to study a selection of statistical approaches, including linear regression, artificial neural networks, and symbolic regression. Our results showed that simpler linear regression models were better able than artificial neural networks to generalize on such a noisy and minimal data set, but there are indications that certain architectures and training configurations can counter overfitting tendencies. This study will be repeated on a more complete SuperCDMS data set (in progress) to explore the interplay between data quality and the application of neural networks.
M. Lupi S. Esposito, X. Llopart-Cudie, A. Pulli
et al.
Verification is a critical aspect of designing front-end (FE) readout ASICs for High-Energy Physics (HEP) experiments. These ASICs share several similar functional features, resulting in similar verification objectives, which can be addressed using comparable verification strategies. This contribution presents a set of re-usable verification components for addressing common verification tasks, such as clock generation, reset handling, configuration, as well as hit and fault injections. The components were developed as part of the CHIPS initiative and they have been successfully used in the verification of multiple HEP ASICs.
W. L. Aldá Júnior, G. A. Alves, K. M. Amarilo
et al.
This proposal outlines the future plans of the Brazilian High-Energy Physics (HEP) community for upcoming collider experiments. With the construction of new particle colliders on the horizon and the ongoing operation of the High-Luminosity LHC, several research groups in Brazil have put forward technical proposals, covering both hardware and software contributions, as part of the Brazilian contribution to the global effort. The primary goal remains to foster a unified effort within the Brazilian HEP community, optimizing resources and expertise to deliver a high-impact contribution to the international HEP community.
Tobias Lammel, P. Boisseaux, M. Fernández-Cruz
et al.
BackgroundGraphene and graphene derivative nanoplatelets represent a new generation of nanomaterials with unique physico-chemical properties and high potential for use in composite materials and biomedical devices. To date little is known about the impact graphene nanomaterials may have on human health in the case of accidental or intentional exposure. The objective of this study was to assess the cytotoxic potential of graphene nanoplatelets with different surface chemistry towards a human hepatoma cell line, Hep G2, and identify the underlying toxicity targets.MethodsGraphene oxide (GO) and carboxyl graphene (CXYG) nanoplatelet suspensions were obtained in water and culture medium. Size frequency distribution of the suspensions was determined by means of dynamic light scattering. Height, lateral dimension and shape of the nanoplatelets were determined using atomic force and electron microscopy. Cytotoxicity of GO and CXYG nanoplatelets was assessed in Hep G2 cells using a battery of assays covering different modes of action including alterations of metabolic activity, plasma membrane integrity and lysosomal function. Induction of oxidative stress was assessed by measuring intracellular reactive oxygen species levels. Interaction with the plasma membrane, internalization and intracellular fate of GO and CXYG nanoplatelets was studied by scanning and transmission electron microscopy.ResultsSupplementing culture medium with serum was essential to obtain stable GO and CXYG suspensions. Both graphene derivatives had high affinity for the plasma membrane and caused structural damage of the latter at concentrations as low as 4 μg/ml. The nanoplatelets penetrated through the membrane into the cytosol, where they were concentrated and enclosed in vesicles. GO and CXYG accumulation in the cytosol was accompanied by an increase in intracellular reactive oxygen species (ROS) levels, alterations in cellular ultrastructure and changes in metabolic activity.ConclusionsGO and CXYG nanoplatelets caused dose- and time-dependent cytotoxicity in Hep G2 cells with plasma membrane damage and induction of oxidative stress being important modes of toxicity. Both graphene derivatives were internalized by Hep G2, a non-phagocytotic cell line. Moreover, they exerted no toxicity when applied at very low concentrations (< 4 μg/ml). GO and CXYG nanoplatelets may therefore represent an attractive material for biomedical applications.