Preface

Timothy Doster, Tegan Emerson, Henry Kvinge, Nina Miolane, Mathilde Papillon, Bastian Rieck, Sophia Sanborn

Abstract: The 2nd ICML Workshop on Topology, Algebra, and Geometry in Machine Learning is an exercise in bringing together researchers working in both of the above threads to exchange ideas, present recent work, and form new collaborations. This proceedings collection captures some of the rich flow of ideas that happened in the workshop. It is also a testament to the breadth of ways that mathematics is currently being applied to modern ML, from the topology of models and datasets to equivariance of models to group actions to applications of hyperbolic geometry to learning problems.