Equivariant Neural Operator Learning with Graphon Convolution

Chaoran Cheng, Jian Peng

We propose a general architecture that combines the coefficient learning scheme with a residual operator layer for learning mappings between continuous functions in the 3D Euclidea…