← Papers

PDE-Refiner

Accurate long rollouts by iterative refinement

Paper title PDE-Refiner: Achieving Accurate Long Rollouts with Neural PDE Solvers

Lippe et al. (Microsoft Research) · General PDEs · Public: code

Paper arXiv 2308.05732PDF

After a normal one-step prediction, the same network refines it a few times by denoising, with noise levels that fall at each step, so that low-amplitude, often high-frequency, parts of the solution are modelled too.

Published
NeurIPS

Share on LinkedIn, X or Bluesky, or or .

Know related work that belongs here?

Something wrong on this page?

More papers

All papers
  1. PhysicsNeMo-Mesh — GPU-Native Meshes and Simulation Cases for Physics AI
  2. Multi-fidelity scaling — Compute-optimal data scaling via multi-fidelity training
  3. CAE-UQ — Predictive Uncertainty for Neural CAE Surrogates