Geometric Clifford Algebra Networks

Explore Clifford Group Equivariant Neural Networks in this comprehensive talk by David Ruhe. Delve into the fundamentals of Clifford algebra and its geometric applications before examining the Clifford group and its action through the orthogonal group. Discover how parameterizations equivariant to the Clifford group automatically become equivariant to the orthogonal group, including rotations and reflections. Learn about polynomial parameterizations under the algebra's geometric product and their significance. Investigate various layers derived from these insights and review experiments conducted in three-, four-, and five-dimensional spaces, including an intriguing application to the nondefinite O(1,3) Lorentz group. Gain practical knowledge on implementing these algorithms using the provided codebase, enhancing your understanding of E(n)-equivariant networks and their applications in geometric algebra.