Neural Harmonic Measure Operator

License: See LICENSE file · Updated 2026-09-28
My rating

This paper introduces the Neural Harmonic Measure Operator, a neural solver for elliptic PDE problems on variable-shape domains. It parameterizes the density of the harmonic measure with a transformer-based boundary kernel supervised by Walk-on-Spheres exit samples. The harmonic measure depends on domain geometry rather than boundary data and can be integrated against boundary data to obtain the Dirichlet Laplace solution.

Tags: