Research map / Probabilistic and scientific modeling / Differential equations and PDEs
Neural Stochastic PDEs: Resolution-Invariant Learning of Continuous Spatiotemporal Dynamics
Cristopher Salvi, Maud Lemercier, Andris Gerasimovics
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 Neural SPDE instead of FNO to learn noise-to-solution map | stochastic KdV · KdV T=0.5, N=1000 | relative L2 error: 0.079 -> 0.004 | Improved |
| use Neural SPDE instead of NCDE-FNO for (u0,xi)->u | stochastic Navier-Stokes · Navier-Stokes | relative L2 error: 0.178 -> 0.049 | Improved |
| use Neural SPDE instead of FNO on noise-to-solution | stochastic Navier-Stokes · Navier-Stokes xi->u | relative L2 error: 0.039 -> 0.034 | Improved |
| increase training set from N=1000 to N=10000 | NSPDE · Ginzburg-Landau (u0,xi)->u | relative L2 error: 0.012 -> 0.006 | Improved |
| drop 10% of observations in time (ablation) | NSPDE · KdV (u0,xi)->u | relative L2 error: 0.008 -> 0.059 | Hurt |
| drop 50% of observations in space (ablation) | NSPDE · KdV xi->u | relative L2 error: 0.004 -> 0.005 | No clear effect |
| observe the driving noise xi in addition to or instead of u0 only | NSPDE · Ginzburg-Landau N=1000 | relative L2 error: 0.128 -> 0.009 | Improved |
| evaluate with fixed-point solver instead of ODE solver | Neural SPDE · runtime | speed: approx. 10 times faster | Improved |
Closest papers by meaning
- Stable Long-Horizon PDE Forecasting via Latent Structured Spectral Propagators (NeurIPS 2026 · similarity 0.90)
- P$^2$C$^2$Net: PDE-Preserved Coarse Correction Network for efficient prediction of spatiotemporal dynamics (NeurIPS 2024 · similarity 0.89)
- SPDEBench: An Extensive Benchmark for Learning Stochastic PDEs (NeurIPS 2026 · similarity 0.89)
- Neural Stochastic Flows: Solver-Free Modelling and Inference for SDE Solutions (NeurIPS 2025 · similarity 0.89)
- Mechanistic PDE Networks for Discovery of Governing Equations (ICML 2025 · similarity 0.89)
- CALM-PDE: Continuous and Adaptive Convolutions for Latent Space Modeling of Time-dependent PDEs (NeurIPS 2025 · similarity 0.89)
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.
