1 paper
Thomas Y. L. Lin, Jiachen Yao, Lufang Chiang +2
We propose a data-efficient, physics-aware generative framework in function space for inverse PDE problems. Existing plug-and-play diffusion posterior samplers represent physics im…