Pith. sign in

REVIEW 13 cited by

MultiScale MeshGraphNets

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2210.00612 v1 pith:ON2PA7K5 submitted 2022-10-02 cs.LG cs.CE

MultiScale MeshGraphNets

classification cs.LG cs.CE
keywords meshgraphnetshigh-resolutionaccuracyaccurategraphimprovingmessagemultiscale
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

In recent years, there has been a growing interest in using machine learning to overcome the high cost of numerical simulation, with some learned models achieving impressive speed-ups over classical solvers whilst maintaining accuracy. However, these methods are usually tested at low-resolution settings, and it remains to be seen whether they can scale to the costly high-resolution simulations that we ultimately want to tackle. In this work, we propose two complementary approaches to improve the framework from MeshGraphNets, which demonstrated accurate predictions in a broad range of physical systems. MeshGraphNets relies on a message passing graph neural network to propagate information, and this structure becomes a limiting factor for high-resolution simulations, as equally distant points in space become further apart in graph space. First, we demonstrate that it is possible to learn accurate surrogate dynamics of a high-resolution system on a much coarser mesh, both removing the message passing bottleneck and improving performance; and second, we introduce a hierarchical approach (MultiScale MeshGraphNets) which passes messages on two different resolutions (fine and coarse), significantly improving the accuracy of MeshGraphNets while requiring less computational resources.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 13 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Physics-Informed Coarsening for Multigrid Graph Neural Surrogates

    cs.LG 2026-05 unverdicted novelty 7.0

    Proposes residual-based physics-informed coarsening in multigrid GNNs to allocate capacity to high-activity regions for more stable solid mechanics surrogates.

  2. Scale-Aware Learning of Chaotic Dynamics on Unstructured Meshes via Binned Spectral Losses

    cs.LG 2026-07 conditional novelty 6.0

    Adding graph-Laplacian band-power matching to the training loss improves long-horizon autoregressive forecasting of chaotic flows on unstructured meshes.

  3. Non-linear mechanical field reconstruction coupling recurrent neural networks with physics-informed graph neural networks

    cs.CE 2026-06 conditional novelty 6.0

    A coupled LSTM-GNN model reconstructs local elasto-plastic stress fields from macroscopic loading paths on a plate-with-hole microstructure, achieving 1000x speedup and mesh transferability with 1.9% error.

  4. PnP-Corrector: A Universal Correction Framework for Coupled Spatiotemporal Forecasting

    cs.AI 2026-05 unverdicted novelty 6.0

    PnP-Corrector decouples physics simulation from error correction to counter reciprocal error amplification in coupled spatiotemporal forecasting, cutting error by 29% in a 300-day ocean-atmosphere test.

  5. PnP-Corrector: A Universal Correction Framework for Coupled Spatiotemporal Forecasting

    cs.AI 2026-05 unverdicted novelty 6.0

    PnP-Corrector decouples physics simulation from error correction via a plug-and-play agent, cutting error by 29% in 300-day global ocean-atmosphere forecasts.

  6. Deep Wave Network for Modeling Multi-Scale Physical Dynamics

    cs.LG 2026-05 unverdicted novelty 6.0

    DW-Net improves the accuracy versus computational cost Pareto front over standard U-Nets for 2D and 3D multi-scale flow benchmarks by stacking multiple waves while keeping training settings identical.

  7. AeTHERON: Autoregressive Topology-aware Heterogeneous Graph Operator Network for Fluid-Structure Interaction

    physics.comp-ph 2026-04 unverdicted novelty 6.0

    AeTHERON achieves mean extrapolation MAE of 0.168 while qualitatively capturing vortex topology on unseen timesteps of flapping flexible caudal fin FSI simulations.

  8. Sparse Autoencoders as a Steering Basis for Phase Synchronization in Graph-Based CFD Surrogates

    cs.CE 2026-03 unverdicted novelty 6.0

    Sparse autoencoders enable phase synchronization in frozen graph CFD surrogates through Hilbert-identified oscillatory features and SVD-based time-varying rotations.

  9. GenDA: Generative Data Assimilation on Complex Urban Areas via Classifier-Free Diffusion Guidance

    cs.LG 2026-01 conditional novelty 6.0

    A sensor-conditioned graph diffusion model reconstructs urban wind fields from <1% of mesh nodes, beating supervised GNN and SVD baselines on held-out altitude slices of one RANS dataset.

  10. 3D Point World Models: Point Completion Enables More Accurate Dynamics Learning

    cs.RO 2026-06 unverdicted novelty 5.0

    3DPWM completes partial point clouds then learns dynamics on the completed 3D scenes to produce reliable long-horizon rollouts for model-based robotic planning.

  11. Mesh Based Simulations with Spatial and Temporal awareness

    cs.LG 2026-05 unverdicted novelty 5.0

    A unified training framework for mesh-based ML surrogates in CFD improves accuracy and long-horizon stability by enforcing spatial derivative consistency via multi-node prediction, using temporal cross-attention corre...

  12. PnP-Corrector: A Universal Correction Framework for Coupled Spatiotemporal Forecasting

    cs.AI 2026-05 unverdicted novelty 4.0

    PnP-Corrector decouples pre-trained physics engines from a correction agent to mitigate reciprocal error amplification in coupled spatiotemporal forecasting, cutting error by 28% on a 300-day ocean-atmosphere task.

  13. GeoTransolver: Learning Physics on Irregular Domains Using Multi-scale Geometry Aware Physics Attention Transformer

    cs.LG 2025-12 conditional novelty 4.0

    GeoTransolver, a geometry-aware attention transformer, improves surrogate CFD accuracy over existing baselines on three automotive/aerospace datasets, but the paper has major reporting gaps.