MENO: Memory-Efficient Neural Operator

MENO (Memory-Efficient Neural Operator) is a neural operator architecture designed for solving Partial Differential Equations, built upon the Manifold Function Encoder framework. It achieves significantly reduced memory consumption with training speed improvements compared to existing architectures, with the notable advantage that memory footprint remains independent of input data resolution. This scalability makes it promising for developing large-scale PDE models across various physics domains.