Jonathan Williams

Jan 17, 2024 | ellis.eu | Jonathan Williams

The new network aims to broaden access to supercomputing, data and software infrastructures, and will operate at the intersection of public and private sectors, academia, and civil society. In addition to ELLIS, ICAIN’s founding members include the Swiss Federal Department of Foreign Affairs, ETH Zurich, the Swiss National Supercomputing Centre, EPFL, Data Science Africa and the Finnish IT Center for Science, which is the hosting member of the LUMI consortium.

Jan 8, 2024 | ellis.eu | Jonathan Williams

Foundation models and their origin, analysis and development have been typically associated with the US and Big Tech. Yet, a critical share of important insights and novel approaches do come from Europe, both within academia and industry. Part of this winter school's goal is to highlight these fresh perspectives and give the students an in-depth look into how Europe is guiding its own research agenda with unique directions and bringing together the community.

Dec 14, 2023 | ellis.eu | Jonathan Williams

ELLIS units connect outstanding AI researchers at a certain location and fulfill a set of strict criteria to ensure scientific excellence. Getting a paper accepted at the NeurIPS conference is an important recognition for any scientist in the field of machine learning. To warm up for the event in New Orleans and to foster exchange among peers, members of 11 ELLIS units in different countries presented their NeurIPS contributions during poster sessions in the week before the conference.

Dec 7, 2023 | ellis.eu | Jonathan Williams

Any European dissertation in the area of artificial intelligence and machine learning-related fields (including computer vision and robotics) can be nominated for the annual award. The dissertation is then reviewed by a committee of renowned scientists for technical depth, the significance of the research contribution and the potential impact on theory and practice. This year, the work of Pepa Atanasova and Robert Geirhos stood out in particular.

Dec 4, 2023 | perceiving-systems.blog | Jonathan Williams

We recently released a pre-print describing PoseGPT, a multi-modal large language model (LLM) that is trained to estimate 3D human pose from images, text, or both. This is interesting because it shows that an LLM can estimate 3D humans from images but this may not be so surprising. After all, LLMs are large capacity models and can be fine-tuned to do many tasks. More interesting and important, is that PoseGPT learns to relate 3D human pose to more general concepts about humans. Figure 1.