PhD position (M/F) - Study of the Higgs boson and its self-coupling in the H → γγ and HH → bbγγ channels using deep learning techniques with the ATLAS experiment at the LHC collider at CERN.

Job Description

Thesis Subject

Particle physics studies the fundamental constituents of matter and their interactions. One of the best tools available to study these particles is the world's most powerful particle accelerator: the Large Hadron Collider (LHC), operated by the European Organization for Nuclear Research (CERN). Among the major advancements achieved with the LHC is the discovery of the Higgs boson by the ATLAS and CMS collaborations in 2012, providing crucial experimental verification of the Standard Model (SM) of particle physics. However, several limitations of the SM remain unexplained. In particular, it does not account for the universe's dark matter content, nor does it include sufficient sources of CP symmetry violation to explain the observed matter-antimatter asymmetry in the universe. The SM is therefore likely a low-energy limit of a more complex theory, and the Higgs field is one of the most useful tools for determining this.

One of the most important experimental studies to undertake is the measurement of the Higgs boson self-coupling. This self-coupling depends on the shape of the Higgs potential, which remains to be verified. The most direct way to measure this self-coupling is to study the production of Higgs boson pairs (di-Higgs), particularly through the HH(bbγγ) channel, which offers a good compromise between the high branching ratio of H(bb) and the high detection sensitivity of the H(γγ) channel. Since the di-Higgs production cross-section is 500 times smaller than that of a single Higgs boson, this process has not yet been detected and could be for the first time with the full dataset from the LHC's Run 3 (by combining different measurement channels). A relatively precise measurement of the self-coupling will only be possible with the data collected during the High-Luminosity phase of the LHC (HL-LHC), starting in 2030.

A significant avenue for improving the analysis is to jointly measure the H(γγ) and HH(bbγγ) channels. In the classical HH(bbγγ) analysis, the H(γγ) channel is treated as a major background process for the di-Higgs channel. A joint analysis of both modes would optimize the overall measurement of the Higgs boson's properties, including notably the Yukawa coupling to the top quark and its self-coupling.

A key component in analyzing the data collected by the ATLAS detector is the reconstruction of the tracks of charged particles produced in the hadron collisions at the LHC. From the measurement of these tracks, fundamental properties of the particles—such as their charge and momentum—can be evaluated. Additionally, we can determine from which interaction vertex they originate, particularly if they come from a displaced vertex, a typical signature of hadrons containing a b quark.

From the start of the HL-LHC, the ATLAS experiment will use a new inner tracker detector, called ITk, which will have greater granularity than the current detector. Reconstructing tracks from the signals collected by the ITk will be a major challenge due to the extremely high combinatorics involved in the high-luminosity conditions of the HL-LHC. Without significant improvements in reconstruction algorithms, the ability to analyze the full dataset collected by the experiment will be severely limited.

The L2IT is a leader in developing new track reconstruction algorithms using geometric deep learning methods, particularly Graph Neural Networks (GNNs), which show great promise. These methods have already demonstrated performance at least comparable to current Track Finding algorithms, with significant room for further development. The next step is to incorporate the evaluation of trajectory parameters (Track Fit) into this process. A promising approach involves regression using Transformers, which have shown excellent performance for this type of task.

Your Work Environment

The thesis will be carried out at the Laboratoire des 2 Infinis – Toulouse (L2IT), within the ATLAS group. The PhD student will join a large-scale international collaboration and participate in the activities of the ATLAS experiment at CERN. The work will focus on the analysis of data from the LHC's Run 3, with prospects toward the High-Luminosity LHC (HL-LHC) program.

Constraints and risks

The project requires a strong proficiency in scientific programming (Python, C++), statistical analysis methods, and large-scale data processing. The PhD student will work in a distributed computing environment (computing grid) and will need to master the software tools developed by the ATLAS collaboration. The work follows the ATLAS collaboration's timeline, with deadlines tied to analyses and publications. Occasional missions to CERN, as well as travel to attend workshops and scientific conferences, are to be expected. No specific risks other than those related to screen work and the use of computer equipment have been identified.

Compensation and benefits

Compensation

2300 € gross monthly

Annual leave and RTT

44 jours

Remote Working practice and compensation

Pratique et indemnisation du TT

Transport

Prise en charge à 75% du coût et forfait mobilité durable jusqu’à 300€

About the offer

About the CNRS

The CNRS is a major player in fundamental research on a global scale. The CNRS is the only French organization active in all scientific fields. Its unique position as a multi-specialist allows it to bring together different disciplines to address the most important challenges of the contemporary world, in connection with the actors of change.

CNRS

The research professions