Join us
Work with us.
If you are excited by machine learning and want to apply it to real problems in collider physics, there is room for you in the group — at Master, PhD, and postdoc level.
Our projects sit at the interface of generative modelling, neural surrogates, and LHC phenomenology. They share a methodological core but differ a lot in flavour: some are mathematically driven, some are engineering- and team-oriented, some are about physics interpretation. We try to match the project to the person.
The descriptions below are intentionally high-level — the concrete work plan is developed together with the supervisor once a direction is chosen, and we are happy to discuss variations or hybrid scopes.
Thesis & project directions
- Open Master thesis
Neural sampling for Monte Carlo integration
A more mathematical project: improving how neural samplers handle difficult regions of phase space. Good for someone who likes thinking about why a method works.
Normalizing flowsPython - Open Master thesis
Uncertainties for neural surrogates
A more team-oriented project: studying how the uncertainty of a neural surrogate carries through to physical observables. A well-defined piece of an ongoing group effort.
StatisticsBasic MLPython - Open Master thesis
Interpretable machine learning for jet tagging
A more physics-oriented project: making jet taggers easier to interpret. Builds on recent group work and suits someone who enjoys connecting ML to physics intuition.
Representation learningPython
How to get in touch
Master students
Email a short note about your background and which direction interests you. No need for a polished proposal — curiosity and some programming experience are the main things.
PhD & postdoc
Positions depend on current funding. Reach out with a CV and a paragraph on your interests; if there is a fit, we will tell you about open calls and fellowships.