Research map / Probabilistic and scientific modeling / Differential equations and PDEs
Neural Operator: Is Data All You Need to Model the World? An Insight Into the Paradigm of Data-Driven Scientific ML
Hrishikesh Viswanath, Md Ashiqur Rahman, Abhijeet Vyas, Andrey Shor, Beatriz Medeiros, Stephanie H. Hernandez, Suhas Eswarappa Prameela, Aniket Bera
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 |
|---|---|---|---|
| replace traditional solver with FNO in Bayesian inversion MCMC | Navier-Stokes Bayesian inversion · Navier-Stokes MCMC | time per computation (s): 2.2 -> 0.005 | Improved |
| use FNO for 30,000 evaluations including training | Navier-Stokes Bayesian inversion · Navier-Stokes MCMC (30,000 samples) | total time: 18 hours -> 12 hours training plus 2.5 minutes | Improved |
| use FNO instead of conventional solver for 100,000 evaluations | AWS p3.2xlarge example · Cost for 100,000 evaluations | cost (USD): 187 -> under 40 | Improved |
| use FNO for few evaluations rather than the conventional solver | FNO surrogate · Small number of evaluations | total compute time: T=43200+0.005n vs T=2.2n | Hurt |
| apply FourCastNet in place of NWP for ensemble forecasting | Global weather forecasting · Ensemble forecasting | speedup: 45,000x speedup | Mixed |
| apply FNO to urban micro-climate modeling | CityFFD · Urban micro-climate | speedup and relative error: 25x speedup, 5% relative error over 1200 time steps | Mixed |
| extend FNO to multivariate dynamics for ocean surrogate | NEMO sea surface height · Coastal flood prediction | speedup: 45x speedup | Improved |
Closest papers by meaning
- Data-Efficient Operator Learning via Unsupervised Pretraining and In-Context Learning (NeurIPS 2024 · similarity 0.87)
- Using Uncertainty Quantification to Characterize and Improve Out-of-Domain Learning for PDEs (ICML 2024 · similarity 0.85)
- Graph Neural PDE Solvers with Conservation and Similarity-Equivariance (ICML 2024 · similarity 0.85)
- Reference Neural Operators: Learning the Smooth Dependence of Solutions of PDEs on Geometric Deformations (ICML 2024 · similarity 0.85)
- RIGNO: A Graph-based Framework For Robust And Accurate Operator Learning For PDEs On Arbitrary Domains (NeurIPS 2025 · similarity 0.84)
- Latent Mamba Operator for Partial Differential Equations (ICML 2025 · similarity 0.84)
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.
