BilAI at the first EurIPS 2025 and the ELLIS UnConference
19.01.2026
Numerous BilAI researchers demonstrated a strong and visible presence at the first EurIPS 2025 in Copenhagen.
EurIPS marks a major milestone for the European AI community: it is a NeurIPS-endorsed European conference featuring exclusively NeurIPS-accepted papers, with the explicit goal of strengthening scientific exchange, visibility, and collaboration within Europe while maintaining the highest international standards in artificial intelligence and machine learning.
BilAI contributions at EurIPS pushing for Sustainability and Security
BilAI played an active role in shaping the scientific agenda of the inaugural EurIPS. Sepp Hochreiter (JKU), Scientific Director of BilAI, delivered a highly visible and well-received Keynote Talk entitled "Sustainable, Low-Energy, and Fast AI Made in Europe". His talk highlighted BilAI’s research vision and underscored Europe’s potential to lead in sustainable and efficient AI.
Further emphasizing its commitment to responsible and trustworthy AI, BilAI sponsored two EurIPS workshops:
BilAI researchers were also represented in the EurIPS poster sessions: Postdoc Edwige Cyffers and PhD candidate Nikita Kalinin, both from ISTA, presented their NeurIPS-accepted posters, contributing cutting-edge research results to the European AI audience.
BilAI at the ELLIS UnConference
In addition to EurIPS, BilAI researchers were actively involved in the ELLIS UnConference, the official kick-off event for EurIPS and a key forum for strengthening the European AI research community.
Christoph Lampert (ISTA), key researcher and member of the BilAI Board of Directors, delivered an invited talk at the ELLIS UnConference workshop “Interactive Learning and Interventional Representations”, where he presented ongoing research from the BilAI cluster.
A major highlight was the first-ever ELLIS workshop on “Energy-Efficient AI: Models, Algorithms, and Hardware for Sustainable Intelligence”.
The workshop was organized by Wolfgang Maass, BilAI key researcher and Director of the ELLIS Unit Graz, together with Marcel van Gerven from the ELLIS Unit Nijmegen. It addressed the rapidly growing energy demands of modern AI systems and the urgent need for sustainable alternatives, including brain-inspired approaches.
BilAI PhD candidate Yukun Yang (TU Graz) shared a new approach to explainable and context-aware reinforcement learning in this workshop. The in-depth discussions among invited experts and junior researchers reflected the growing momentum in this field, pushing for more support for this line of research in Europe, e.g. through an ELLIS Research Program on energy-efficient AI.
