Research map / Image generation / Fast diffusion sampling
Solving Inverse Physics Problems with Score Matching
Benjamin Holzschuh, Simona Vegetti, Nils Thuerey
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 multi-step training instead of 1-step training (ablation) | SMDP · 1D Toy SDE | posterior metric Q | Improved |
| Increase the maximum sliding window size (ablation) | SMDP · Stochastic heat equation | reconstruction MSE | Improved |
| Use SDE inference instead of ODE inference (ablation) | SMDP · Stochastic heat equation | spectral loss | Improved |
| Use ODE inference instead of SDE inference (ablation) | SMDP · Stochastic heat equation | reconstruction MSE | Improved |
| Remove the reverse physics step (ablation) | SMDP · Stochastic heat equation | spectral loss | Hurt |
| Set the score multiplier C to 1 instead of 2 for SDE inference (ablation) | SMDP · Buoyancy-driven flow with obstacles | performance | Improved |
| Add the learned score correction to the reverse physics surrogate (ablation) | SMDP · Navier-Stokes with unknown dynamics | reconstruction MSE | Improved |
| Use SDE inference instead of ODE inference (ablation) | SMDP · Navier-Stokes with unknown dynamics | spectral error | Improved |
| Use ODE inference instead of SDE inference (ablation) | SMDP · Navier-Stokes with unknown dynamics | reconstruction MSE | Improved |
Closest papers by meaning
- InverseBench: Benchmarking Plug-and-Play Diffusion Priors for Inverse Problems in Physical Sciences (ICLR 2025 · similarity 0.88)
- Improving Diffusion-based Inverse Algorithms under Few-Step Constraint via Linear Extrapolation (NeurIPS 2025 · similarity 0.87)
- Self-diffusion for Solving Inverse Problems (NeurIPS 2025 · similarity 0.87)
- Repulsive Latent Score Distillation for Solving Inverse Problems (ICLR 2025 · similarity 0.86)
- DMPlug: A Plug-in Method for Solving Inverse Problems with Diffusion Models (NeurIPS 2024 · similarity 0.86)
- Physics-Informed Deep Inverse Operator Networks for Solving PDE Inverse Problems (ICLR 2025 · 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.
