Research map / Probabilistic and scientific modeling / Differential equations and PDEs
Conditionally Parameterized, Discretization-Aware Neural Networks for Mesh-Based Modeling of Physical Systems
Jiayang Xu, Aniruddhe Pradhan, Karthikeyan Duraisamy
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 first layer with CP-Dense layer conditioned on convection and diffusion terms | CNN closure model · Burgers, training IC | u-bar avg MAE: 0.23 -> 0.15 | Improved |
| same, final step | CNN closure model · Burgers, testing IC | u-bar final MAE: 0.23 -> 0.13 | Improved |
| replace first layer with CP-Dense layer | DDP closure model · Burgers, training IC | u-bar avg MAE: Inf. -> 0.42 | Improved |
| replace first layer with CP-Dense layer | DDP closure model · Burgers, testing IC | u-bar final MAE: Inf. -> 0.41 | Improved |
| replace second dense layer with CP-Dense conditioned on log local Reynolds number | MLP super-resolution · channel flow, training | Ex avg abs error: 0.0145 -> 0.0120 | Improved |
| same | MLP super-resolution · channel flow, testing | Ey max abs error: 0.0675 -> 0.0519 | Improved |
| same | MLP super-resolution · channel flow, testing | Ex avg abs error: 0.0098 -> 0.0081 | Improved |
| replace dense layers with CP layers (CP-GNet) | GNet · rocket injector rollout | prediction accuracy | Improved |
| use smaller CP-GNet5L (1.3M params) instead of GNet15L (1.8M) | GNet · rocket injector rollout | prediction accuracy: 1.8M -> 1.3M parameters | Improved |
| train CP-GNet10 on short period instead of long (ablation) | CP-GNet · rocket injector rollout | prediction accuracy | No clear effect |
Closest papers by meaning
- Parameterized Physics-informed Neural Networks for Parameterized PDEs (ICML 2024 · similarity 0.89)
- UGM2N: An Unsupervised and Generalizable Mesh Movement Network via M-Uniform Loss (NeurIPS 2025 · similarity 0.88)
- PIG: Physics-Informed Gaussians as Adaptive Parametric Mesh Representations (ICLR 2025 · similarity 0.87)
- P$^2$C$^2$Net: PDE-Preserved Coarse Correction Network for efficient prediction of spatiotemporal dynamics (NeurIPS 2024 · similarity 0.87)
- Graph Neural PDE Solvers with Conservation and Similarity-Equivariance (ICML 2024 · similarity 0.86)
- Vectorized Conditional Neural Fields: A Framework for Solving Time-dependent Parametric Partial Differential Equations (ICML 2024 · similarity 0.86)
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.
