Research map / Probabilistic and scientific modeling / Differential equations and PDEs
Solving Nonlinear PDEs with Sparse Radial Basis Function Networks
Zihan Shao, Konstantin Pieper, Xiaochuan Tian
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 |
|---|---|---|---|
| use randomly generated instead of grid collocation points (ablation) | Sparse RBFNet and GP · Semilinear Poisson | accuracy: performance degrades with random collocation points | Hurt |
| increase number of collocation points K from 400 to 2500 (random points) (ablation) | Sparse RBFNet alpha=1e-4 · Semilinear Poisson | L2 error: 7.30e-1 -> 3.88e-2 | Improved |
| use Sparse RBFNet instead of Gaussian Process (random points, K=2500) | Sparse RBFNet · Semilinear Poisson | L2 error: 4.97e-1 -> 8.89e-2 | Improved |
| use Sparse RBFNet instead of Gaussian Process (random points, K=400) | Sparse RBFNet · Semilinear Poisson | L2 error: 4.79e-1 vs 7.16e-1 | Mixed |
| decrease regularization alpha from 1e-2 to 1e-4 (random points, K=2500) (ablation) | Sparse RBFNet · Semilinear Poisson | L2 error: 8.89e-2 -> 3.88e-2 | Improved |
| increase K from 40 to 200 for Burgers with nu=0.002 (ablation) | Sparse RBFNet · Burgers equation (nu=0.002) | solution quality: solution improves significantly from K=40 to 200 | Improved |
| increase K from 200 to 400 for Burgers with nu=0.002 (ablation) | Sparse RBFNet · Burgers equation (nu=0.002) | solution quality: additional gain relatively modest | No clear effect |
Closest papers by meaning
- Toward Efficient Kernel-Based Solvers for Nonlinear PDEs (ICML 2025 · similarity 0.86)
- Towards General Neural Surrogate Solvers with Specialized Neural Accelerators (ICML 2024 · similarity 0.86)
- Global convergence of adjoint-optimized neural PDEs (NeurIPS 2026 · similarity 0.86)
- Mechanistic PDE Networks for Discovery of Governing Equations (ICML 2025 · similarity 0.86)
- PINNacle: A Comprehensive Benchmark of Physics-Informed Neural Networks for Solving PDEs (NeurIPS 2024 · similarity 0.85)
- Physics-Informed Neural Network Policy Iteration: Algorithms, Convergence, and Verification (ICML 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.
