The workshop brings together researchers working at the interface of physics, machine learning, and the emerging physics of intelligence across a remarkably broad range of scales, from elementary particles and quantum many-body systems to condensed matter, nuclear physics, astrophysics, and cosmology.
Recent advances in machine learning are opening new ways to extract physical information from complex data, discover hidden structures, accelerate numerical simulations, and formulate effective descriptions of strongly interacting systems. At the same time, ideas and methods from physics are increasingly being used to understand the principles underlying learning, representation, inference, and collective behavior in complex systems. This two-way interaction, using machine learning to advance physics, and using physics to deepen our understanding of intelligence, is a central theme of the conference.
Hosted by the Institute for Physics of Intelligence (IPI) at The University of Tokyo, this international conference aims to foster interactions among communities that do not usually meet and to explore common concepts, methods, and challenges that transcend conventional disciplinary boundaries. By bringing together researchers from diverse areas of physics, mathematics, and machine learning, we hope to uncover new connections across scales and contribute to a broader physical understanding of both natural phenomena and intelligent systems.