Neural PDE Solving on Complex Obstacles
0.0001MSEDGNet
Evaluation results
| Method | Links | ||
|---|---|---|---|
| DGNetApproach=discrete Green's function embedding2026.03 | 0.0001 | 0.0211 | |
| DGNetTraining Regime=limited-trajectory2026.03 | 0.0001 | 0.0211 | |
| TransolverArchitecture=Transformer architecture2026.03 | 0.0009 | 0.1855 | |
| AMG (Harnessing)Architecture=graph-based architecture2026.03 | 0.0014 | 0.3058 | |
| MP-PDETraining Regime=limited-trajectory2026.03 | 0.0061 | 0.541 | |
| MGNTraining Regime=limited-trajectory2026.03 | 0.0078 | 0.612 | |
| SINGERArchitecture=multi-scale architecture2026.03 | 0.0094 | 0.7746 | |
| DeepONetTraining Regime=limited-trajectory2026.03 | 0.0533 | 0.5061 | |
| BENOTraining Regime=limited-trajectory2026.03 | 0.0766 | 0.6069 | |
| PhyMPGNTraining Regime=limited-trajectory2026.03 | 0.297 | 1.1956 |