Hugo Simon
hugo.simon🎩cea.fr
I am a PhD candidate in Cosmology and Machine Learning at CEA Paris-Saclay, supervised by Arnaud de Mattia and François Lanusse. I am also a member of the DESI collaboration, which is building the largest 3D map of the Universe to date.
I am particularly interested in field-level modeling and scalable Bayesian inference, leveraging fast and differentiable cosmological models and high-dimensional ($\geq 10^6$) sampling methods, to extract maximum information from cosmological surveys. My current efforts center on field-level inference of primordial non-Gaussianity from DESI.
selected publications
- JCAPBenchmarking field-level cosmological inference from galaxy redshift surveys2025Accepted to JCAP
- NeurIPSSmall coresets via negative dependence: DPPs, linear statistics, and concentrationNeurIPS 2024 - Thirty-eigth Conference on Neural Information Processing Systems, 2024Accepted as Spotlight Paper. Authors are listed in alphabetical order