Research map / Probabilistic and scientific modeling / Differential equations and PDEs
Uncertainty-Aware Diagnostics for Physics-Informed Machine Learning
Mara Daniels, Liam Hodgkinson, Michael W. Mahoney
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 |
|---|---|---|---|
| select kernel bandwidth by minimizing the PILE score (ablation) | PIKL (Gaussian process physics-informed kernel learning) · 2D Poisson equation, 13x13 noisy observations | fit quality (PPL2-G data and physics error) | Improved |
| tune physics and data regularization (rho, gamma) sequentially by minimizing PILE (ablation) | PIKL · 2D Poisson equation | PILE score / PPL2-G error | Improved |
| too small or too large kernel bandwidth instead of the PILE-optimal one (ablation) | PIKL · 2D Poisson equation | derivative fit | Hurt |
| use isotropic RBF kernel with bandwidth selection by PILE (ablation) | PIKL · 1D convection PDE | data and physics loss | Hurt |
| select anisotropic RBF kernel parameters by minimizing the data-free PILE (Fredholm determinant) (ablation) | PIKL with RBF kernel · 1D convection PDE | data and physics fit | Improved |
Closest papers by meaning
- Uncertainty Quantification for Physics-Informed Neural Networks with Extended Fiducial Inference (NeurIPS 2025 · similarity 0.87)
- Understanding Generalization in Physics Informed Models through Affine Variety Dimensions (NeurIPS 2025 · similarity 0.87)
- Calibrated Physics-Informed Uncertainty Quantification (ICML 2025 · similarity 0.86)
- Physics and Lie symmetry informed Gaussian processes (ICML 2024 · similarity 0.86)
- Neural Oscillators for Generalization of Physics-Informed Machine Learning (AAAI 2024 · similarity 0.85)
- S$^{2}$-PINN: Stochastic Separable Physics-Informed Neural Networks (NeurIPS 2026 · 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.
