Optimal Convergence Rates for Neural Operators

Mike Nguyen, Nicole Mücke

We introduce the neural tangent kernel (NTK) regime for two-layer neural operators and analyze their generalization properties. For early-stopped gradient descent (GD), we derive f…