Fractional Laplace Neural Operators: Exact Architectures, an Expressivity Frontier at Criticality, and Certified Stability for Memory-Driven Network Dynamics
My rating
The resource addresses memory-driven hereditary network dynamics represented through Volterra resolvents with non-rational Laplace symbols. Its fractional Laplace neural operator embeds this structure into the learned map, using one block graph-spectral layer to represent the full linear Volterra solution operator exactly for commuting excitation–Laplacian pairs. It also establishes an expressivity frontier for finite rational realizations and studies certified stability.
Tags: