Erlangen AI Hub hosted the second day of its conference at Oxford Mathematics, showcasing AI research topics including autoformalization, feature learning, AI reasoning, scientific discovery, online safety regulation, hierarchical data, and grokking.

The Erlangen AI Hub uses powerful tools from mathematical and algorithmic fields, including geometry, topology, and category theory, to deepen our understanding of AI systems and advance the development of safer, more reliable next-generation AI. Inspired by Felix Klein’s Erlangen Programme, which formalised geometry through the notions of symmetry and invariance expressed through group theory, we pursue foundational approaches to modern AI research. Our collaborative ecosystem brings together leading expertise from across the UK, bridging academia, industry, and public discourse on AI policy and governance.