arXiv subjects

Rachik Soualah

Publications and source records attributed to Rachik Soualah.

10 recordsLinked to original sources

arXiv2026

Multitask Jet Analysis with Vision-Language Models: A Physics-Informed Four-Panel Representation

The enormous recorded data at high-energy physics (HEP) colliders make the accurate identification of physics objects a bottleneck in disentangling event topologies, where machine learning has become a standard tool for jet tagging. In this work, we examine whether Vision-Language Models (VLMs) can provide a common interface for structured jet analysis from a single physics-informed image. Each JetClass jet becomes a 224x224 RGB image with four panels encoding p_T flow of all constituents; charged-hadron, neutral-hadron, and electromagnetic composition; p_T-weighted impact-parameter significances with displaced-track multiplicity; and signed-track p_T densities with local p_T^k-weighted jet-charge asymmetry. Four open VLMs are adapted with low-rank adaptation (LoRA) for class classification across QCD, Higgs, W/Z, and top jets, six-field attribute prediction, and cross-panel consistency with replaced-panel localization, evaluated on 24000 balanced JetClass test jets. Ablations identify the impact-parameter lifetime signature of heavy-flavor tagging as dominant, with jet charge separating the nearly mass-degenerate hadronic W and Z. Zero-shot recall stays near chance (3.33%-10.57%), once adapted, Task 1 Macro recall grows monotonically with training size, and Gemma4-E4B reaches 75.05% (74.87% F_1), 93.31% Task 2 field-mean, and 99.88%/99.75% binary/localization recall. Reallocating budget toward W/Z jets raises Zqq recall by up to 9.83 points at near-constant Macro recall. Transfer to broader-topology JetClass-II and real-data Aspen Open Jets (probing the simulation-to-data gap) is retained: with 3000 target jets, Task 1 recall exceeds 70% on JetClass-II and every Task 2/3 metric exceeds 92%. Thus rasterizing energy flow, species, displacement, and jet charge, with parameter-efficient adaptation, supports jet analysis through one instruction-conditioned interface for future collider data.

hep-ph↗

arXiv2025

Image and Point-cloud Classification for Jet Analysis in High-Energy Physics: A survey

Nowadays, there has been a growing trend in the field of high-energy physics (HEP), in both its experimental and phenomenological studies, to incorporate machine learning (ML) and its specialized branch, deep learning (DL). This review paper provides a thorough illustration of these applications using different ML and DL approaches. The first part of the paper examines the basics of various particle physics types and establishes guidelines for assessing particle physics alongside the available learning models. Next, a detailed classification is provided for representing Jets that are reconstructed in high-energy collisions, mainly in proton-proton collisions at well-defined beam energies. This section covers various datasets, preprocessing techniques, and feature extraction and selection methods. The presented techniques can be applied to future hadron-hadron colliders (HHC), such as the high-luminosity LHC (HL-LHC) and the future circular collider - hadron-hadron (FCChh). The authors then explore several AI techniques analyses designed specifically for both image and point-cloud (PC) data in HEP. Additionally, a closer look is taken at the classification associated with Jet tagging in hadron collisions. In this review, various state-of-the-art (SOTA) techniques in ML and DL are examined, with a focus on their implications for HEP demands. More precisely, this discussion addresses various applications in extensive detail, such as Jet tagging, Jet tracking, particle classification, and more. The review concludes with an analysis of the current state of HEP using DL methodologies. It highlights the challenges and potential areas for future research, which are illustrated for each application.

hep-ph↗

arXiv2025

Resummed jet mass distribution with trimming in Z+jet events at the LHC

In this paper, we calculate the resummed jet mass distribution up to next-to-leading logarithmic accuracy for jets defined with the trimming groomer in Z+jet events at the LHC. We compute at fixed order and to all orders the large logarithms, both in the jet mass and $z_{\mathrm{cut}}$ trimming variable, including non-global and clustering logarithms. We conduct phenomenological studies at $\mathcal{O}(α_s)$, validating our results by comparing them with the fixed-order Monte Carlo program MCFM, which provides NLO predictions for hadron colliders. The resummation of non-global logarithms is estimated in the large-$\mathrm{N_c}$ limit. Our resummed result is compared to Monte Carlo parton shower simulations from Pythia 8, Herwig ++ and Sherpa 2 event generators.

hep-ph↗

arXiv2021

The Scale Invariant Scotogenic Model: Dark Matter and the Scalar Sector

In this paper, we investigate the mutual impact between the dark matter (DM) requirements and the scalar sector in the scale invariant (SI) scotogenic model. The model is motivated by the neutrino mass and DM within a classically SI framework. It is a SI generalization of the scotogenic model, where the standard model (SM) is extended by a real singlet, an inert scalar doublet and three Majorana singlet fermions, where the lightest one ($N_{1}$) could play the DM candidate role. In addition to the annihilation channels $N_{1}N_{1}\rightarrow\ell_α\ell_β,ν_α\barν_β$, the DM can be annihilated via few s-channel processes into SM particles, that are mediated by the Higgs/dilaton. This allows the new Yukawa interactions, that are responsible for neutrino mass generation, to take small values and therefore avoid the mass degeneracy between the CP-even and CP-odd inert scalars unlike the case of the minimal scotogenic model. In contrast to many Majorana DM models, the DM in the SI-scotogenic model couples to the quarks at tree-level, and hence the constraint from the direct detection experiments is very important on the space parameter. The aim of this work is to investigate the correlation between the DM requirements and the scalar sector in this model.

hep-ph↗

arXiv2021

Searching for GeV-scale Majorana Dark Matter: inter spem et metum

We suggest a minimal model for GeV-scale Majorana Dark Matter (DM) coupled to the standard model lepton sector via a charged scalar singlet. We show that there is an anti-correlation between the spin-independent DM-Nucleus scattering cross-section ($σ_{\mathrm{SI}}$) and the DM relic density for parameters values allowed by various theoretical and experimental constraints. Moreover, we find that even when DM couplings are of order unity, $σ_{\mathrm{SI}}$ is below the current experimental bound but above the neutrino floor. Furthermore, we show that the considered model can be probed at high-energy lepton colliders using e.g. the mono-Higgs production and same-sign charged Higgs pair production.

hep-ph↗

arXiv2019

Probing the Majorana Nature in Radiative Neutrino mass models with the same-sign dilepton final states at future colliders

Neutrinos in the Standard Model (SM) are considered to be massless which is in contradiction with the evidence from the neutrino oscillation data. These experiments established that at least two SM neutrinos have non-zero masses and that individual lepton numbers are violated. This is strong evidence of new physics beyond the SM that should be responsible for generating non zero mass for the neutrinos. In this work, we study the collider phenomenology of an extension of the SM where neutrinos are generated radiatively at three-loop. We show that the production of same-sign dilepton at lepton colliders (such as FCC-ee and ILC) can be used to probe the Majorana nature of neutrinos in this class of models.

hep-ph↗

arXiv2016

Probing Radiative Neutrino Mass Models With Dilepton Events At The LHC

In this work we prob a class of neutrino mass models at both Large Hadron Collider (LHC) energies 8 TeV and 14 TeV. The focus will be on the new introduced interaction terms between a singlet charged scalar, $S^{\pm}$, and leptons leading to different final states $pp\rightarrow\ell_α^{\pm}\ell_β^{\mp}$ +$\slashed E$ with $\ell_α\ell_β=ee,eμ,μμ$ that implies lepton flavor violation (LFV). An accurate cut on the $M_{T2}$ eventvariable is found to be crucial for an effective suppression of the large Standard Model background. The obtained results can be translated into a possible detectability of the charged scalars effect..

hep-ph↗

arXiv2016

Probing Radiative Neutrino Mass Models with Dilepton Events at the LHC

In this work, we investigate the possibility of probing a class of neutrino mass models at the LHC proton-proton collisions with 8 and 14 TeV energies. The existence of lepton flavor violating interactions for a singlet charged scalar, $S^{\pm}$, that couples to the leptons could induce many processes such as $pp\rightarrow\ell_α^{\pm}\ell_β^{\mp}+\slashed E $. Using the processes with $\ell_α\ell_β=ee,eμ,μμ$, we found that an inclusive cut on the $M_{T2}$ event variable is vital in our analysis and leads to an effective suppression of the large Standard Model background. Our results show possible detectability of the charged scalars effect, especially at the $\sqrt{s}= 14~\text{TeV}$.

hep-ph↗

arXiv2014

Radiative Neutrino Mass Model at the $e^{-}e^{+}$ Linear Collider

We study the phenomenology of a Standard Model (SM) extension with two charged singlet scalars and three right handed (RH) neutrinos at an electron-positron collider. In this model, the neutrino mass is generated radiatively at three-loop, the lightest RH neutrino is a good dark matter candidate; and the electroweak phase transition strongly first order as required for baryogenesis. We focus on the process $e^{+}+e^{-}\rightarrow e^{-}μ^{+}+E_{miss}$, where the model contains new lepton flavor violating interactions that contribute to the missing energy. We investigate the feasibility of detecting this process at future $e^{-}e^{+}$ linear colliders at different center of mass energies: $E_{CM}$=250, 350, 500 GeV and 1 TeV.

hep-ph↗

arXiv2012

Measurement of the charge asymmetry in top quark pair production in pp collisions at $\sqrt s = 7 $ TeV using the ATLAS detector

We present a measurement of the charge asymmetry in top-antitop production using data corresponding to an integrated luminosity of $0.70 fb^{-1}$ of proton-proton collisions at $\sqrt s = 7$ TeV collected by the ATLAS detector. The top pair events decaying semileptonically (lepton+jets channel) to either an electron or muon, missing transverse energy and at least four jets are selected. The reconstruction of the $t\bar{t}$ events was performed using a kinematic likelihood approach. The difference of absolute values of top and antitop rapidities is used to define the charge asymmetry: $A_{C} = (N(|ΔY|>0) - N(|ΔY|<0)) / (N(|ΔY>0) + N(|ΔY|<0))$. To allow comparisons with theory calculations, a Bayesian unfolding technique is applied to correct the measured $|ΔY|$ distributions for acceptance and detector effects. The top charge asymmetry in both channels (e and mu) after correction is measured to be: $A_{C} = -0.009 \pm 0.023 (stat) \pm 0.032 (syst)$ (e+jets channel) and $A_{C} = -0.028 \pm 0.019 (stat) \pm 0.022 (syst)$ ($μ$+jets channel) giving a combined result of : $A_{C} = -0.024 \pm 0.016 (stat) \pm 0.023 (syst)$. These results are compatible with the Standard Model predictions of $A_{C}=0.006$.

hep-ex↗