SCOPE: Observation-Conditioned Full-Target Prediction for Sparse PDE Inference

License: See LICENSE file · Updated 2026-09-29
My rating

SCOPE is a research method for recovering complete PDE fields from sparse observations. It combines full-field latent prediction with physical reconstruction to infer states that are not uniquely determined by limited measurements. The paper contrasts this approach with diffusion-based PDE solvers that use iterative sampling and neural operators that make deterministic one-pass predictions.

Tags: