TRANSFORMER: from the genealogy of dark matter halos to the baryonic properties of galaxy clusters.
The thesis proposes to predict the baryonic properties of galaxy clusters based on the history of dark matter halo formation, using innovative neural networks (Transformers). The work will involve intensive numerical simulations. This project falls within the general framework of determining cosmological parameters through the observation of galaxy clusters in X-rays. It is directly linked to the international Heritage programme in the XMM-Euclid FornaX deep field.
Study of radiative decay of the nucleus using a technic like Oslo-method
Radiative neutron capture is a nuclear reaction forming a compound nucleus which decays by emitting gamma-rays at excitation energy around the neutron binding energy. This well-known reaction which we known how to accurately measure its cross section at low incident neutron energies for most stable and few unstable nuclei close the stability valley, remains difficult to measure for exotic nuclei like fission fragments. Nuclear reaction models based essentially on stable nuclei, also struggle to provide reliable predictions of cross sections for these exotic nuclei. However, in the recent years, progress made related to the models and the measurements for the radiative capture show that significant improvements in including microscopic ingredients studies. These micoscopic ingredients: gamma strength function and nuclear level density, remain accesible to the experiment. These ingredients which respectively manage the way of how the gamma cascade occurs and the nuclear structure at high excitation energy can also be measured and calculated to be compared and suggest ways to improve the predictability of models. This kind of improvements have a direct impact for instance on the cross sections for these exotic nuclei which are produced in the stellar nucleosynthesis. The subject of thie thesis is to measure these quantities for a nucleus involved in the nucleosythesis using a new setup called SFyNCS.
Machine Learning-Based Algorithms for Real-Time Standalone Tracking in the Upstream Pixel Detector at LHCb
This PhD aims to develop and optimize next-generation track reconstruction capabilities for the LHCb experiment at the Large Hadron Collider (LHC) through the exploration of advanced machine learning (ML) algorithms. The newly installed Upstream Pixel (UP) detector, located upstream of the LHCb magnet, will play a crucial role from Run 5 onward by rapidly identifying track candidates and reducing fake tracks at the earliest stages of reconstruction, particularly in high-occupancy environments.
Achieving fast and highly efficient tracking is essential to fulfill LHCb’s rich physics program, which spans rare decays, CP-violation studies in the Standard Model, and the characterization of the quark–gluon plasma in nucleus–nucleus collisions. However, the increasing event rates and data complexity expected for future data-taking phases will impose major constraints on current tracking algorithms, especially in heavy-ion collisions where thousands of charged particles may be produced per event.
In this context, we will investigate modern ML-based approaches for standalone tracking in the UP detector. Successful applications in the LHCb VELO tracking system already demonstrate the potential of such methods. In particular, Graph Neural Networks (GNNs) are a promising solution for exploiting the geometric correlations between detector hits, allowing for improved tracking efficiency and fake-rate suppression, while maintaining scalability at high multiplicity.
The PhD program will first focus on the development of a realistic GEANT4 simulation of the UP detector to generate ML-suitable datasets and study detector performance. The next phase will consist in designing, training, and benchmarking advanced ML algorithms for standalone tracking, followed by their optimization for real-time GPU-based execution within the Allen trigger and reconstruction framework. The most efficient solutions will be integrated and validated inside the official LHCb software stack, ensuring compatibility with existing data pipelines and direct applicability to Run-5 operation.
Overall, the thesis will provide a major contribution to the real-time reconstruction performance of LHCb, preparing the experiment for the challenges of future high-luminosity and heavy-ion running.
Impact of irradiation parameters on the alpha’ phase formation in oxide dispersion strengthened steels
Ferritic-martensitic oxide dispersion strengthened steels (ODS steels) are materials of great interest in the nuclear industry. Predominantly composed of iron and chromium, these materials can become brittle due to the precipitation of a chromium-rich phase, called a', under irradiation. This phase, known to be sensitive to irradiation conditions, provides an ideal topic for a deeper exploration of the capability to emulate neutron irradiation with ions. Indeed, while ion irradiations are frequently used to understand phenomena observed during neutron irradiations, the question of their representativeness is often raised.
In this thesis, we aim to understand how the irradiation parameters can affect the characteristics of the a' phase in ODS steels. To do so, various ODS steels will be irradiated under different conditions (flux, dose, temperature, and type of particles, such as ions, neutrons, electrons), and subsequently analyzed at the nanoscale. The a' phase (size, chromium content) obtained for each ion irradiation condition will be compared to the one after neutron irradiation.
INVESTIGATION OF THE NUCLEAR TWO-PHOTON DECAY
The nuclear two-photon, or double-gamma decay is a rare decay mode in atomic nuclei whereby a nucleus in an excited state emits two gamma rays simultaneously. This second-order electromagnetic process, well known in atomic physics, has been little studied for the atomic nucleus due to the largely predominant first-order processes. Even-even nuclei with a first excited 0+ state are favorable cases to search for a double-gamma decay branch, since the emission of a single gamma ray is strictly forbidden for 0+ to 0+ transitions by angular momentum conservation. The double-gamma decay still remains a very small decay branch (<1E-4) competing with the dominant (first-order) decay modes of atomic internal-conversion electrons (ICE) or internal positron-electron (e+-e-) pair creation (IPC).
The thesis project has two distinct experimental parts: First, we store bare (fully-stripped) ions in their excited 0+ state in the heavy-ion storage ring (ESR) at the GSI facility to search for the double-gamma decay in several nuclides. For neutral atoms the excited 0+ state is a rather short-lived isomeric state with a lifetime of the order of a few tens to hundreds of nanoseconds. At relativistic energies available at GSI, however, all ions are fully stripped of their atomic electrons and decay by ICE emission is hence not possible. If the state of interest is located below the pair creation threshold the IPC process is not possible either. Consequently, bare nuclei are trapped in a long-lived isomeric state, which can only decay by double-gamma emission to the ground state. The decay of the isomers is identified by so-called time-resolved Schottky Mass Spectroscopy. This method allows to distinguish the isomer and the ground state by their (very slightly) different revolution time in the ESR, and to observe the disappearance of the isomer peak in the mass spectrum with a characteristic decay time. Successful experiment establishing the double-gamma decay in several nuclides (72Ge, 98Mo, 98Zr) were already performed and a new experiment to study the nuclide 194Pb has been accepted by the GSI Programme Committee and its realization is planned for 2027.
The second part concerns the direct observation of the emitted photons using gamma-ray spectroscopy. While the storage ring experiments allow to measure the partial lifetime for the double gamma decay, further information on the nuclear properties can be only be achieved by measuring the photon themselves. A test experiment has been performed to study its feasibility and the plans a more detailed study should be developed with the PhD project.
Joint simulation-based inference of tSZ maps and Euclid's weak lensing
Context:
The Euclid mission will provide weak lensing measurements with unprecedented precision, which have the potential to revolutionise our understanding of the Universe. However, as the statistical uncertainties decrease, controlling systematic effects becomes even more crucial. Among these, baryonic feedback, which redistributes gas within galaxies and clusters, remains one of the key astrophysical systematic effects limiting Euclid’s ability to constrain the equation of state of dark energy. Understanding baryonic feedback is one of the urgent challenges of cosmology today.
The thermal Sunyaev-Zel’dovich (tSZ) effect provides a unique window into the baryonic component of the Universe. This effect arises from the scattering of cosmic microwave background (CMB) photons by hot electrons in galaxy groups and clusters. This is the same hot gas that has been redistributed by baryonic feedback and is particularly relevant for weak lensing cosmology. The cross-correlation between tSZ and weak lensing (WL) probes how baryons trace and modify the cosmic structures, allowing joint constraints on cosmology and baryonic physics.
Most current tSZ-WL analyses rely on fitting angular power spectra under the assumption of a Gaussian likelihood. However, the tSZ signal is highly non-Gaussian, as it traces the massive structures of the Universe, and the power spectra fail to fully capture the information in the data. To unlock the scientific potential of the tSZ-WL analyses, it is essential to move beyond these simplifying assumptions.
PhD thesis:
The goal of this PhD project is to develop a novel simulation-based framework to jointly analyse tSZ and Euclid’s WL data. This framework will combine physically motivated forward models with advanced statistical and machine-learning techniques to provide accurate measurements of baryonic feedback and cosmological parameters. By jointly analysing tSZ and WL measurements, this project will increase the accuracy of Euclid’s cosmological analyses and improve our understanding of the dark matter-baryon connection.
Contribution of artificial intelligence to the study of fission
Nuclear fission is an extreme process during which a heavy nucleus deforms until it reaches a point of no return leading to its separation into two fragments. The process goes with a significant release of energy, mainly as kinetic energy of the newly formed fragments, but also as excitation energy (about 15 MeV/fragment). In addition, the fragments are also produced with a high angular momentum. It is through the emission of neutrons and photons that fission fragments evacuate their energy and angular momentum. The ultimate experiment in fission would consist of identifying each fragment in mass and charge; measuring their kinetic energy; and characterize in energy and multiplicity the neutrons and photons they emit. This data set would make it possible to access the global energy of the fission process and to completely characterize the deexcitation of the fragments. Due to the significant complexity of such an exclusive measurement, this data set is always missing.
Our team is moving towards such measurement and this thesis work aims to explore the benefits that machine learning techniques can bring in this perspective.
The thesis will consist of taking advantage of all the experimentally accessible multi-correlated data in order to feed machine learning algorithms whose purpose will be to identify fission fragments and determine their properties.
The developed techniques will be applied to a first data set using a twin ionization chamber for the detection of fission fragments coupled to a set of neutron detectors. The data will be acquired at the beginning of the thesis.
In a second step, a more exploratory study will consist of applying the same techniques to data obtained during the thesis using a temporal projection chamber as a fission fragment detector. It will be a matter of demonstrating that the energy resolution is compatible with the study of fission.
Real-time measurement of edge plasma parameters for the optimization of the WEST tokamak performance
The control of heat fluxes at the plasma edge, and particularly in the divertor — a dedicated volume where these fluxes are focused — is a major challenge in research on magnetic confinement fusion. In future devices, heat fluxes will need to be dissipated by radiation to reduce the heat conducted to the divertor. However, the operational window for these high-radiation regimes is quite narrow and requires precise control of the edge plasma. The PhD first objective is to develop a real-time measurement of the density and temperature profiles at the plasma edge from the Thomson scattering diagnostics. By leveraging a large experimental database and simulations performed with edge plasma modeling tools and plasma/wall interaction models, the student will then develop meta-models to create a real-time control algorithm for WEST scenarios, particularly for high radiation discharges. This development will rely on continuous iteration between simulations, experimental observations, and real-time control performance requirements. This thesis is part of a collaborative framework involving French universities and international collaborations, with a high level of expected scientific visibility.
Machine-learning methods for the cosmological analysis of weak- gravitational lensing images from the Euclid satellite
Weak gravitational lensing, the distortion of the images of high-redshift galaxies due to foreground matter structures on large scales, is one of the most promising tools of cosmology to probe the dark sector of the Universe. The statistical analysis of lensing distortions can reveal the dark-matter distribution on large scales, The European space satellite Euclid will measure cosmological parameters to unprecedented accuracy. To achieve this ambitious goal, a number of sources of systematic errors have to be quanti?ed and understood. One of the main origins of bias is related to the detection of galaxies. There is a strong dependence on local number density and whether the galaxy's light emission overlaps with nearby objects. If not handled correctly, such ``blended`` galaxies will strongly bias any subsequent measurement of weak-lensing image distortions.
The goal of this PhD is to quantify and correct weak-lensing detection biases, in particular due to blending. To that end, modern machine- and deep-learning algorithms, including auto-differentiation techniques, will be used. Those techniques allow for a very efficient estimation of the sensitivity of biases to galaxy and survey properties without the need to create a vast number of simulations. The student will carry out cosmological parameter inference of Euclid weak-lensing data. Bias corrections developed during this thesis will be included a prior in galaxy shape measurements, or a posterior as nuisance parameters. This will lead to measurements of cosmological parameters with a reliability and robustness required for precision cosmology.
Control of trapped electron mode turbulence with an electron cyclotron resonant source
The performance of a tokamak-type fusion power plant in term of energy gain will be limited by turbulent transport. The instability of trapped electron modes is one of the main instabilities causing turbulence in tokamaks. Furthermore, electron cyclotron resonance heating (ECRH) is the generic heating system in current and future tokamaks. Both physical processes are based on resonant interactions with electrons, in space and velocity. Since heating has the effect of depopulating the resonant interaction zone of its electrons, superimposing its resonance on that of the instability can theoretically lead to a stabilisation of the trapped electron modes.
The objective of the thesis is twofold: (i) to construct scenarios where this mechanism exists and validate it using linear simulations, then (ii) to characterise its effect and quantify its effectiveness in non-linear regimes where linear effects compete with the self-organisation of turbulence, with collisional processes and with the dynamics of average profiles. Potentially, this entirely new control technique could improve the performance of tokamaks at no additional cost. The PhD thesis will require a detailed theoretical understanding of the two resonant processes and their various control parameters. It will be based on the use of the high performance computing gyrokinetic code GYSELA dedicated to the study of transport and turbulence in tokamak plasmas, which has recently been enhanced with an ECRH heating module. An experimental component is also planned on the WEST and/or TCV tokamaks to validate the identified most promising turbulence control scenario(s).