QuadNorm: Resolution-Robust Normalization for Neural Operators
About
Normalization layers in neural operators usually compute statistics by uniformly averaging discrete grid values, making the normalization itself discretization-dependent and thereby a source of transfer error across different resolutions or meshes. To enable discretization robustness, we introduce a quadrature normalization family that replaces existing uniform averaging in normalization layers with numerical quadrature: QuadNorm and BlendQuadNorm. On endpoint-inclusive uniform grids, the proposed quadrature moments are $O(h^2)$-consistent across discretizations, meaning that their cross-resolution mismatch decays quadratically with grid spacing. A transfer-error bound then predicts how normalization-induced mismatch scales with both the resolution gap and network depth. The experiments show the same gap- and depth-scaling trends predicted by the transfer-error bound. On Darcy, QuadNorm delivers the best cross-resolution performance at every tested target resolution from $64^2$ to $256^2$; on real-data benchmarks, Transolver with QuadNorm achieves nearly resolution-invariant transfer. The largest gains appear on nonperiodic PDEs and nonspectral architectures, where native-resolution improvements also emerge. We also validate BlendQuadNorm, which stays close to LayerNorm behavior and serves as a conservative default for periodic FNO settings. These results identify normalization as a previously overlooked source of resolution dependence in neural operators.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| PDE solving | Darcy | Relative L2 Error3.78 | 46 | |
| Forward PDE solving | Elasticity | Relative L2 Error9.72 | 44 | |
| PDE solving | Darcy Flow 64x64 resolution | Relative L2 Error0.0561 | 27 | |
| PDE solving | Darcy Flow 128x128 resolution | Relative L2 Error6.19 | 17 | |
| PDE solving | Darcy Flow 256x256 resolution | Relative L2 Error6.66 | 17 | |
| PDE solving | Darcy Flow 32x32 resolution | Relative L2 Error0.0327 | 17 | |
| Fluid Flow Prediction | Darcy Flow 700 train and 200 test examples official FNO (test) | Relative L2 Error6.32 | 15 | |
| PDE solving | FNO Darcy 64² → 128² resolution official (test) | Relative L2 Error3.08 | 12 | |
| PDE solving | FNO Darcy 64² → 256² resolution official (test) | Relative L2 Error3.4 | 12 | |
| PDE solving | Cavity flow | Relative L2 Error1.3 | 12 |