UFO: A Domain-Unification-Free Operator Framework for Generalized Operator Learning

Introduces UFO, a cross-domain neural operator framework enabling operators through adaptive interactions among representations in distinct domains (physical, spectral, latent), achieving discretization decoupling where input functions can be observed at variable resolutions.