Operator Learning on PDEBench CNS (η=1, ζ=0.1)
0.0043L2REAOT-POT-S
Evaluation Results
| Method | Links | |
|---|---|---|
| AOT-POT-SParams=31M, Evaluation Protocol=Fine-tuned2026.05 | 0.0043 | |
| AOT-POT-TParams=7M, Evaluation Protocol=Fine-tuned2026.05 | 0.0049 | |
| AOT-POT-MParams=124M, Evaluation Protocol=Fine-tuned2026.05 | 0.0051 | |
| AOT-POT-SParams=31M, Evaluation Protocol=Zero-shot2026.05 | 0.0077 | |
| AOT-POT-MParams=124M, Evaluation Protocol=Zero-shot2026.05 | 0.0091 | |
| MoE-POT-MParams=288M, Evaluation Protocol=Fine-tuned2026.05 | 0.0096 | |
| DPOT-MParams=122M, Evaluation Protocol=Fine-tuned2026.05 | 0.01 | |
| MoE-POT-TParams=17M, Evaluation Protocol=Fine-tuned2026.05 | 0.0108 | |
| AOT-POT-TParams=7M, Evaluation Protocol=Zero-shot2026.05 | 0.011 | |
| MoE-POT-MParams=288M, Evaluation Protocol=Zero-shot2026.05 | 0.0112 | |
| DPOT-TParams=7M, Evaluation Protocol=Fine-tuned2026.05 | 0.0112 | |
| DPOT-MParams=122M, Evaluation Protocol=Zero-shot2026.05 | 0.0116 | |
| MoE-POT-SParams=90M, Evaluation Protocol=Fine-tuned2026.05 | 0.0122 | |
| DPOT-SParams=30M, Evaluation Protocol=Fine-tuned2026.05 | 0.0129 | |
| MoE-POT-SParams=90M, Evaluation Protocol=Zero-shot2026.05 | 0.0143 | |
| DPOT-SParams=30M, Evaluation Protocol=Zero-shot2026.05 | 0.0153 | |
| MoE-POT-TParams=17M, Evaluation Protocol=Zero-shot2026.05 | 0.016 | |
| DPOT-TParams=7M, Evaluation Protocol=Zero-shot2026.05 | 0.0173 | |
| FFNOParams=1.3M, Evaluation Protocol=Individually trained2026.05 | 0.0212 | |
| GNOTParams=1.8M, Evaluation Protocol=Individually trained2026.05 | 0.0325 | |
| GK-TParams=1.6M, Evaluation Protocol=Individually trained2026.05 | 0.0341 | |
| OformerParams=1.9M, Evaluation Protocol=Individually trained2026.05 | 0.0417 | |
| FNOParams=0.5M, Evaluation Protocol=Individually trained2026.05 | 0.098 | |
| FNO-mParams=7M, Evaluation Protocol=Zero-shot2026.05 | 0.151 | |
| UNetParams=25M, Evaluation Protocol=Individually trained2026.05 | 0.334 |