Short Bio

Sebastian is Full Professor at the Research Institute for Statistics and Information Science at the University of Geneva. His research group works on:

• AI weather forecasting
• Extreme value theory and graphical models
• Extrapolation in machine learning
• Statistical climate science

Sebastian did his studies in Mathematics at University of Göttingen and UC Berkeley, and he obtained his PhD in 2013 at the University of Göttingen. He was then an SNSF Ambizione fellow at EPF Lausanne with Anthony Davison, and visiting professor at the Department of Statistical Sciences at the University of Toronto from 2018–2019. From 2020-2026 he was holding an SNSF Eccellenza grant.

Memberships and Responsibilities

• Co-Chair of Climate Informatics 2026 Conference taking place in Lausanne, April 27-30, 2026.
• Member of the Executive Committee of the Bernoulli Society as Membership Secretary (2019 – 2025).
• Member of the Research Section Committee (editorial board for JRSSB discussion papers) (2024 – ).
• Associate Editor of JRSSB (2021 – 2025), Extremes (2018 – ) and Scandinavian Journal of Statistics (2019 – 2022).
• Chair of the Scientific Committee for the Extreme Value Analysis Conference 2025 in Chapel Hill.
• Director of the Western Swiss Doctoral School in Statistics and Probability CUSO (2020 – 2023).
• Member of ISI (elected), Bernoulli Society, RSS, ASA.

News

• May 2026: I am thrilled to participate as PI in the NCCR CLIM+ on climate extremes, a large interdisciplinary project recently been granted by the SNSF!
• May 2026: Join us at EGU26: we are convening a session on AI for Weather and Climate: Opportunities and Challenges
• April 2026: Join us at the Climate Informatics 2026 Conference taking place in Lausanne, April 27-30, 2026. Registration is open! We also host a hackathon sponsored by Nvidia!
• March 2026: Our paper showing that "Numerical models outperform AI weather forecasts of record-breaking extremes" (with Z. Zhang, E. Fischer, J. Zscheischler) has been accepted in Science Advances.
• February 2026: New paper on "A Kullback-Leibler divergence test for multivariate extremes: theory and practice" (with P. Naveau and C. Zhou) on arxiv.
• January 2026: Our paper on "Boosted control functions" (with N. Gnecco, J. Peters and N. Pfister) has been accepted to the Journal of Machine Learning Research.
• January 2026: New paper on "Graph structure learning for stable processes" (with F. Brück and S. Volgushev) on arxiv.
• December 2025: New paper on "Intrinsic Whittle-Matérn fields and sparse spatial extremes" (with D. Bolin, P. Braunsteins and R. Huser) on arxiv.
• November 2025: Our paper on "Learning extremal graphical structures in high dimensions." (with M. Lalancette and S. Volgushev) has been accepted to the Annals of Statistics.
• September 2025: I gave an interviews on our work on AI weather models for this Bloomberg article by Joe Wertz and for the Tagesanzeiger article by Joachim Laukenmann.
• August 2025: New paper on "Numerical models outperform AI weather forecasts of record-breaking extremes" (with Z. Zhang, E. Fischer and J. Zscheichler) on arxiv.
• August 2025: I am now Full Professor at the Research Institute for Statistics and Information Science at University of Geneva !
• June 2025: New paper on "Theoretical guarantees for neural estimators in parametric statistics" (with A. Rödder and M. Hentschel) on arxiv.
• May 2025: Happy to be cited in this New York Times article where I comment on the new Microsoft AI foundational weather model Aurora that just appeared in Nature.
• May 2025: Our paper "Graphical models for infinite measures with applications to extremes" (with J. Ivanovs and K. Strokorb) been accepted in Annals of Applied Probability.
• March 2025: New paper on "Extreme Conformal Prediction: Reliable Intervals for High-Impact Events" (with O. Pasche and H. Lam) on arxiv.
• March 2025: New paper on "Extremes of structural causal models" (with N. Gnecco and F. Röttger) on arxiv.
• February 2025: Our paper on "Extremal graphical modeling with latent variables via convex optimization" (with A. Taeb) has been accepted in JMLR.
• October 2024: New paper on "Progression: an extrapolation principle for regression" (with G. Buriticá) on arxiv.
• October 2024: New paper on "Lévy graphical models" (with J. Ivanovs and J. D. Thøstesen) on arxiv.
• September 2024: I am honored to serve as the Chair of the Scientific Committee for the Extreme Value Analysis Conference 2025 taking place in Chapel Hill from June 23 to June 27, 2025!!
• June 2024: Our paper on "Statistical Inference for Hüsler-Reiss Graphical Models Through Matrix Completions" (with M. Hentschel and J. Segers) has been accepted in JASA, Theory and Methods. [arxiv]
• April 2024: Our paper on "Neural networks for extreme quantile regression with an application to forecasting of flood risk" (with O. Pasche) has been accepted in Annals of Applied Statistics. [arxiv]
• April 2024: New paper on "Validating Deep-Learning Weather Forecast Models on Recent High-Impact Extreme Events" (with O. Pasche, J. Wider, Z. Zhang and J. Zscheichler) on arxiv.
• March 2024: New paper on "Extremal graphical modeling with latent variables" (with A. Taeb) on arxiv.
• February 2024: We hosted a workshop and mini-courses on 'Causality in Extremes' at University of Geneva. Many thanks to everyone for making this such an interactive event!
• Feburary 2024: New review article on "Graphical models for multivariate extremes" (with M. Hentschel, M. Lalancette and F. Roettger) on arxiv.