Research map / Probabilistic and scientific modeling / Differential equations and PDEs
Generalized Spherical Neural Operators: Green’s Function Formulation
Hao Tang, Hao Chen, Chao Li
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add GSNO blocks and multi-scale network in place of SFNO | SFNONet · SSWE 256x512, 5h | average MRE (x10^-3): 147 -> 136 | Improved |
| add GSNO blocks and multi-scale network in place of SFNO | SFNONet · SSWE 256x512, 10h | average MRE (x10^-3): 195 -> 176 | Improved |
| add GSNO block in place of SFNO block (ablation) | SFNONet · WeatherBench, 5 days | average ACC (%): 49.2 -> 51.3 | Improved |
| add GSNO block in place of SFNO block (ablation) | GSNO network · WeatherBench, 5 days | average ACC (%): 50.9 -> 52.5 | Improved |
| add GSNO block in place of SFNO block (ablation) | GSNO network · WeatherBench, 1 day | average ACC (%): 89.9 -> 91.5 | Improved |
| add multi-scale U-Net structure with spherical up-down sampling (ablation) | SFNO with GSNO blocks · WeatherBench, 3 days | average ACC (%): 70.5 -> 72.4 | Improved |
| use GSNO spectral kernel instead of spatial position-embedding SNO | SNO · WeatherBench operator-level, 120h | ACC (%): 50.3 -> 52.5 | Improved |
| increase SFNO channels from 64 to 96 (ablation) | SFNO · WeatherBench operator-level, 120h | ACC (%): 50.9 -> 50.7 | No clear effect |
| add correction term I(T_corr) to GSNO (ablation) | GSNO · WeatherBench, day 5 | average ACC (%): 42.6 -> 52.5 | Improved |
| add original term I(T_orig) to GSNO (ablation) | GSNO · WeatherBench, day 5 | average ACC (%): 9.9 -> 52.5 | Improved |
| use GSNO in place of SFNO in FOD-Net architecture | FOD-Net · HCP whole brain | ACC: 0.8334 -> 0.8517 | Improved |
| use GSNO instead of SFNO at lower resolution 64x128 | SFNO · WeatherBench 64x128, 5 days | ACC: 49.2 -> 51.7 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ClimaX — ClimaX: A foundation model for weather and climate (ICML 2023) · better on 4 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Neural Green’s Functions (NeurIPS 2025 · similarity 0.88)
- Beyond Regular Grids: Fourier-Based Neural Operators on Arbitrary Domains (ICML 2024 · similarity 0.86)
- Neural Operators with Localized Integral and Differential Kernels (ICML 2024 · similarity 0.86)
- Learning semilinear neural operators: A unified recursive framework for prediction and data assimilation. (ICLR 2024 · similarity 0.86)
- RIGNO: A Graph-based Framework For Robust And Accurate Operator Learning For PDEs On Arbitrary Domains (NeurIPS 2025 · similarity 0.85)
- Optimal deep learning of holomorphic operators between Banach spaces (NeurIPS 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
