HEIBRiDS - Helmholtz Einstein International Berlin Research School in Data Science

Big things today: the Helmholtz-ELLIS Foundation Models in Science workshop is here, and Helmholtz AI is thrilled to be part of it! The workshop brings together leading minds exploring?how foundation models are reshaping scientific discovery - from?next-gen large language models (LLMs) to AI-driven breakthroughs in physics, health, and materials science. ?? Day 1:?Cutting-edge research in?LLMs and multimodal AI - the very technologies defining the future of machine learning. ?? Day 2:?Real-world impact - how AI is revolutionizing scientific discovery across disciplines. With experts from?Meta, Microsoft, ETH Zurich, Simons Foundation, ELLIS, and beyond, this is where?AI meets the frontiers of science. We can’t wait for all the insightful talks, thought-provoking discussions, and new ideas that will shape the next era of research! ?? About the workshop: https://lnkd.in/dgeeBAht ?? About the Helmholtz Foundation Model Initiative: https://lnkd.in/dwEfHjSH The Helmholtz-ELLIS Foundation Models in Science workshop is organized by the Helmholtz-Gemeinschaft / Helmholtz Association Foundation Model Initiative – Synergy Unit & the ELLIS - European Laboratory for Learning and Intelligent Systems (ELLIS) and is supported by Helmholtz AI, Helmholtz Imaging, HEIBRiDS - Helmholtz Einstein International Berlin Research School in Data Science & Zuse School ELIZA.

Exciting times ahead! Helmholtz Imaging is thrilled to be part of the HFMI-ELLIS Workshop on Foundation Models in Science, happening today & tomorrow. This workshop brings together top minds at the intersection of AI and scientific research, exploring large language models (LLMs) and multimodal AI, and showcasing AI’s revolutionary applications in specific scientific fields. Helmholtz Imaging will kick off both days with welcoming remarks from: ? Dagmar Kainmueller, Head of Research Unit at Max Delbrück Center – Day 1 ? Fabian Isensee, Head of Engineering and Support Unit at DKFZ German Cancer Research Center, together with Bernhard Sch?lkopf from ELLIS – Day 2 We’re looking forward to inspiring talks, engaging discussions, and fresh perspectives! If you’re attending, let’s connect and exchange ideas. ?? Don’t miss Lena Maier-Hein, one of Helmholtz Imaging’s Center Coordinators, presenting “Rethinking Foundation Model Evaluation” on Day 1 in the afternoon! More information about this workshop ?? bit.ly/HFMI-ELLIS-WS This workshop is organized by the Helmholtz-Gemeinschaft / Helmholtz Association Foundation Model Initiative – Synergy Unit & the ELLIS - European Laboratory for Learning and Intelligent Systems (ELLIS), and supported by Helmholtz AI, #HelmholtzImaging, HEIBRiDS - Helmholtz Einstein International Berlin Research School in Data Science & Zuse School ELIZA. #Helmholtz #HelmholtzImaging #HFMI #FoundationModels #AI #LLM

Last HEIBRiDS Lecture for this semester, where Grégoire Montavon from #BIFOLD presented his "Insights into Unsupervised Learning Models with Explainable AI". See you back in the summer semester with a new series of #HEIBRiDS Lectures.

????? Join the BIFOLD Talk with #CharlottePelletier: Exploring Advances in Satellite Image Time Series Analysis. Feb. 13, '25, 2pm. #TUBerlin, EN148. We are pleased to invite researchers and professionals working at the intersection of remote sensing and AI to the upcoming BIFOLD Talk, featuring Charlotte Pelletier. ?? Date: Febuary 13, 2025 ?? Location: Technische Universit?t Berlin, Room EN148, Einsteinufer 17, 10587 Berlin High-resolution satellite imagery plays a critical role in monitoring terrestrial surfaces over time. However, the vast volume and complexity of Satellite Image Time Series (#SITS) introduce challenges that require advanced AI-driven solutions. In this talk, Charlotte Pelletier will discuss: ?? Graph and covariance matrix-based approaches for SITS classification ?? A novel super-resolution strategy for generating high-resolution imagery ?? Domain adaptation techniques for transferring models to new regions This session will be particularly relevant for AI researchers, remote sensing specialists, and geospatial data scientists seeking insights into the latest developments in time-series analysis. We look forward to welcoming you onsite for an engaging discussion. #RemoteSensing #AIforEarth #SatelliteImaging #GeospatialData #MachineLearning #TimeSeriesAnalysis #BIFOLDTalk Begüm Demir

Congratulations Hermann!

Congratulations Daniel and team!

The first HEIBRiDS lecture of the year took place yesterday at MDC-BIMSB, where Nadja Klein, Head of the Methods for Big Data Lab at the Karlsruhe Institute of Technology (KIT), presented "A Statistical Perspective on Bayesian Deep Learning." #heibrids #heibridslectureseries

Congratulations Tabea Rettelbach ! We will miss your enthusiasm and energy in #HEIBRIDS

Yesterday, I had the pleasure of participating in a podium discussion at a festive event hosted by HEIBRiDS. HEIBRiDS - Helmholtz Einstein International Berlin Research School in Data Science, is a graduate program for data science students. It brings together six Helmholtz Centers and four university partners, including the Einstein Center Digital Future and BIFOLD - Berlin Institute for the Foundations of Learning and Data, offering fully-funded PhD projects at the intersection of Data Management, Machine/Deep Learning, Imaging, Mathematical Modelling, and High-throughput Data Analytics. On the podium, Prof. Nadja Klein, Prof. Wojciech Samek, and I discussed “the role of data science in the changing scientific and political landscape.” Significant changes are underway, such as the further internationalization of the data sciences, the importance of interdisciplinarity in addressing pressing issues of our time, and the role of science-industry collaboration in tackling these challenges. We've discussed this with depth. Thanks to HEIBRiDS and the Einstein Center Digital Future for organizing the event, and special thanks to Elke Witt for moderating the discussion. #EinsteinCenterDigitalFuture #ECDF #DataScience #HEIBRiDS #BiFOLD Sandra Pravica Samira Franzel

Five BIFOLD research teams are hiring: two PostDoc position and three PhD positions focused on machine learning, data management, and their intersection. Check our opportunity website. Link in the comments. #AI #research #jobs #PhDCandidates #PostDoc #MachineLearing #DataManagement #PhD