You are cordially invited to the fourth CESOC Members’ Assembly taking place at Forschungszentrum Jülich (interactive site plan of FZJ) – home of Europe’s fastest computer: JUPITER.
.
Highlights:
– Exclusive tour of the JUPITER Exascale computer—you’ll get to see Europe’s fastest supercomputer up close
– Keynotes by Professors Anke Noelscher, director of the Troposphere Division (ICE-3) at the Institute of Climate and Energy, and Aleksandar Bojchevski, leading the Trustworthy Artificial Intelligence Lab at the University of Cologne (see below)
– Election of six board directors – your vote counts!
Feel free to bring colleagues; we would be delighted to welcome new CESOC members https://cesoc.net/become-a-member/.
Travel Arrangements (Mitfahrangelegenheiten): If you are driving or would like a ride, please get in touch with info@cesoc.net
Please, note: you are only allowed to access the campus if you have registered with CESOC in advance. A valid form of identification (ID card, passport, or driver’s license) is required for entry.
We look forward to a productive and hopefully enjoyable gathering in Jülich!
Anke Noelscher on “Variability in space and time(s): About the art of uncovering statistical pattern in observational atmospheric data”
Abstract:
Atmospheric composition determines local air quality, impacts weather and governs global climate. Variations in space bridge from micro to global scales. Variations in time picture entangled transport and transformation processes. This talk discusses current challenges and chances for working on observational atmospheric data using novel statistical approaches.
Bio:
Anke Nölscher is an atmospheric scientist exploring Earth-atmosphere interactions. She is particularly interested in the specific organic molecular composition of air that can be like a fingerprint, revealing information about its origin, chemistry, transport and fate, and potential impact on human and environmental health. Anke Nölscher obtained her PhD at the Max Planck Institute for Chemistry in Mainz in 2012, joined the California Institute of Technology in Pasadena (US) in 2015 as a Postdoc, and worked as a project coordinator at the German Weather Services (DWD). She tenured at the University of Bayreuth and became Director of the Institute of Climate and Energy (ICE-3, Forschungszentrum Jülich) and Professor for Atmospheric Chemistry at the University of Cologne in 2025.
Aleksandar Bojchevski on “Taming Unpredictable AI”
Abstract:
Artificial Intelligence is everywhere, including high-stakes domains where mistakes can be disastrous. The problem? These models can be unpredictable and easy to trick. So how can we trust them? First, we’ll show just how easily AI models can be fooled. Then, we’ll introduce techniques that wrap unreliable black-box models in strong statistical guarantees, making their predictions more trustworthy.
Bio:
Aleksandar Bojchevski is a Professor of Computer Science at the University of Cologne, where he leads the Trustworthy Artificial Intelligence Lab (TAIL). His research is about making AI trustworthy and efficient, by building models that are safe, reliable, and, where possible, come with provable guarantees about their behaviour.