Pith. sign in

REVIEW 53 cited by

Learning Mesh-Based Simulation with Graph Networks

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

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 2010.03409 v4 pith:3IQKLXNQ submitted 2020-10-07 cs.LG cs.CE

Tobias Pfaff , Meire Fortunato , Alvaro Sanchez-Gonzalez , Peter W. Battaglia This is my paper · ORCID
classification cs.LGcs.CE
keywords complexgraphlearningmeshmesh-basedsimulationsimulationsdynamics
verification ladder T0 review T1 audit T2 compute T3 formal

Signed reviews

No signed human review yet.

0 comments
read the original abstract

Mesh-based simulations are central to modeling complex physical systems in many disciplines across science and engineering. Mesh representations support powerful numerical integration methods and their resolution can be adapted to strike favorable trade-offs between accuracy and efficiency. However, high-dimensional scientific simulations are very expensive to run, and solvers and parameters must often be tuned individually to each system studied. Here we introduce MeshGraphNets, a framework for learning mesh-based simulations using graph neural networks. Our model can be trained to pass messages on a mesh graph and to adapt the mesh discretization during forward simulation. Our results show it can accurately predict the dynamics of a wide range of physical systems, including aerodynamics, structural mechanics, and cloth. The model's adaptivity supports learning resolution-independent dynamics and can scale to more complex state spaces at test time. Our method is also highly efficient, running 1-2 orders of magnitude faster than the simulation on which it is trained. Our approach broadens the range of problems on which neural network simulators can operate and promises to improve the efficiency of complex, scientific modeling tasks.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 53 Pith papers

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

  1. FourCastNet 3: A geometric approach to probabilistic machine-learning weather forecasting at scale

    cs.LG 2025-07 conditional novelty 7.0 of 10

    A purely convolutional, spherical-geometry weather model trained with a combined spatial and spectral CRPS loss delivers GenCast-level skill, IFS-beating accuracy, and stable spectra out to 60 days.

  2. GReFEM: Multimodal LLMs as Zero-Shot Semantic Assistants for Physics-Guided 3D Mesh Refinement

    cs.GR 2026-07 conditional novelty 6.5 of 10

    GReFEM shows MLLMs zero-shot isolate load-activated geometric features for volumetric mesh refinement with higher precision than matched-budget geometric heuristics.

  3. Label-Free Finite-Volume-Residual Training of Attention Graph Neural Networks for Coupled Thermo-Fluid Fields

    physics.flu-dyn 2026-07 conditional novelty 6.0 of 10

    An attention graph neural network trained only against finite-volume residuals predicts coupled 3D thermo-fluid fields without labeled CFD data, matching or beating a supervised baseline on four benchmarks.

  4. Image Editing Models are Numerical Solvers

    cs.CV 2026-07 conditional novelty 6.0 of 10

    A FLUX image editor, adapted per task, maps rendered inputs to rendered solutions across ten numerical problems, but several encodings leak the target's scale and the model cannot track chaotic Kuramoto-Sivashinsky dynamics.

  5. A Research Prototype for Closed-Loop Generative Design of Customized Foot Orthoses via Semantic-Physics Alignment

    cs.AI 2026-07 conditional novelty 6.0 of 10

    TANS-FO maps clinical text to lattice insole geometry and uses a GNN stress surrogate in a closed loop, reporting R²=0.94 vs FEA and a 34.7% surrogate-predicted peak-pressure reduction over parametric CAD.

  6. Spatio-Temporal Prediction of Unsteady Airfoil Aerodynamics Using Augmented Graph Neural Ordinary Differential Equations with Exogenous Controls

    cs.LG 2026-07 conditional novelty 6.0 of 10

    GNODE, a graph-neural-network-plus-neural-ODE surrogate with augmented latent dimensions, predicts unsteady transonic airfoil flows with more stable and accurate rollouts than an autoregressive GNN baseline.

  7. Evaluation of State-of-the-Art Deep Learning Architectures for Aerodynamical Predictions

    physics.flu-dyn 2026-07 conditional novelty 6.0 of 10

    Benchmarking four neural operators for airfoil and NASA CRM pressure prediction: Transolver best on 2D, BSMS-GNN best on 3D; UPT and GAOT lag.

  8. A multi-scale feature enhanced graph neural network for fluid dynamics prediction in complex geometries

    cs.LG 2026-07 conditional novelty 6.0 of 10

    ME-GNN reaches SOTA relative L2 errors of 0.0196/0.0556 on ShapeNet-Car velocity/pressure, 0.0033 NMSE on AirfRANS, and 0.1416 on DrivAerNet surface pressure by combining multi-scale U-Net features with local two-step...

  9. BubbleSH: A Dataset of Rising Bubbles with Deformable Interfaces

    cs.LG 2026-07 conditional novelty 6.0 of 10

    A lightweight DNS-derived dataset of deformable bubble swarms with spherical-harmonic shape descriptors enables data-driven emulators that achieve 10^4-10^5x speedup over direct simulation.

  10. RynnWorld-4D: 4D Embodied World Models for Robotic Manipulation

    cs.RO 2026-07 conditional novelty 6.0 of 10

    A tri-branch diffusion model co-generates RGB, depth, and optical flow from a single RGB-D image, and an inverse dynamics head on its internal latents achieves state-of-the-art bimanual manipulation success rates.

  11. Quadrature-Aware Complex-Linear Neural Operator for Boundary-to-Field Prediction in Resonant Acoustics

    physics.flu-dyn 2026-07 conditional novelty 6.0 of 10

    A quadrature-aware complex-linear neural operator halves field error versus DeepONet and enforces exact source superposition for resonant cavity acoustics.

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

    cs.LG 2026-07 conditional novelty 6.0 of 10

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

  13. Adaptive Physics Transformer with Fused Global-Local Attention for Subsurface Energy Systems

    cs.LG 2026-02 conditional novelty 6.0 of 10

    APT, a mesh-agnostic neural operator fusing graph-based local features with global attention, is claimed to be the first architecture trained directly on adaptive-mesh-refinement simulations and outperforms state-of-t...

  14. A hybrid global local computational framework for ship hull structural analysis using homogenized model and graph neural network

    cs.CE 2025-12 conditional novelty 6.0 of 10

    A coarse equivalent-single-layer hull model feeds boundary conditions to a graph transformer that predicts local stiffened-panel stress and displacement fields.

  15. Spatio-temporal, multi-field deep learning of shock propagation in meso-structured media

    cs.LG 2025-09 conditional novelty 6.0 of 10

    A CNN-LSTM surrogate autoregressively predicts seven coupled shock fields in meso-structured materials with 1.4-3.2% RMSE, 94% better than single-field models.

  16. M4GN: Mesh-based Multi-segment Hierarchical Graph Network for Dynamic Simulations

    cs.LG 2025-09 conditional novelty 6.0 of 10

    A hierarchical mesh-graph network with modal-decomposition-guided segmentation reports up to 56% lower rollout error than baselines and introduces a long-range beam-deformation benchmark.

  17. TRELLIS-Enhanced Surface Features for Comprehensive Intracranial Aneurysm Analysis

    cs.CV 2025-09 conditional novelty 6.0 of 10

    TRELLIS-derived surface features improve aneurysm classification, segmentation, and hemodynamic simulation, including a 15% lower blood-flow prediction error.

  18. Hybrid Physics-Machine Learning Models for Quantitative Electron Diffraction Refinements

    physics.comp-ph 2025-08 unverdicted novelty 6.0 of 10

    A differentiable physics-machine-learning hybrid jointly fits structural parameters and neural-network corrections for experimental effects in 3D electron diffraction refinement.

  19. Graph Neural Network Surrogates for Contacting Deformable Bodies with Necessary and Sufficient Contact Detection

    cs.CE 2025-07 conditional novelty 6.0 of 10

    A graph neural network surrogate with continuous, necessary-and-sufficient contact detection and a contact-loss regularizer improves generalization for soft-body contact prediction and achieves up to about a thousand-...

  20. Parallel Data Object Creation: Towards Scalable Metadata Management in High-Performance I/O Library

    cs.DC 2025-06 conditional novelty 6.0 of 10

    A new netCDF header format with an index table and independent metadata blocks enables scalable parallel creation of millions of data objects, cutting creation time by up to 582x at 4,096 processes.

  21. Dynamical Data for More Efficient and Generalizable Learning: A Case Study in Disordered Elastic Networks

    physics.chem-ph 2025-05 conditional novelty 6.0 of 10

    A graph neural network trained only on compression trajectories of disordered elastic networks infers Poisson's ratio and generalizes to networks with Poisson's ratios outside the training range.

  22. Thickness-aware E(3)-Equivariant 3D Mesh Neural Networks

    cs.CV 2025-05 conditional novelty 6.0 of 10

    A mesh GNN that links opposite surfaces via learned thickness edges improves node-level 3D deformation prediction while a PCA-based canonical coordinate system preserves E(3) equivariance.

  23. Towards scalable surrogate models based on Neural Fields for large scale aerodynamic simulations

    physics.flu-dyn 2025-05 conditional novelty 6.0 of 10

    MARIO, a modulated conditional neural field with learned SDF geometry codes, predicts RANS flow fields and surface pressures with reported order-of-magnitude accuracy gains on the AirfRANS scarce task and strong resul...

  24. DISCO: learning to DISCover an evolution Operator for multi-physics-agnostic prediction

    cs.LG 2025-04 conditional novelty 6.0 of 10

    A hypernetwork reads a short trajectory and outputs the parameters of a small neural ODE-like PDE solver, achieving state-of-the-art next-frame prediction on PDEBench with significantly fewer training epochs.

  25. Predicting Stress and Damage in Carbon Fiber-Reinforced Composites Deformation Process using Composite U-Net Surrogate Model

    cs.LG 2025-04 conditional novelty 6.0 of 10

    Composite-Net, an auto-regressive composite U-Net surrogate, predicts the full stress and damage field evolution in a 2D carbon fiber composite under uniaxial strain, with under 20 MPa stress RMSE on all tested cases ...

  26. Rigid Body Adversarial Attacks

    cs.CV 2025-02 conditional novelty 6.0 of 10

    The authors optimize internal material distributions to create objects indistinguishable in rigid-body simulation but with dramatically different deformable-simulation trajectories.

  27. Factorized Implicit Global Convolution for Automotive Computational Fluid Dynamics Prediction

    cs.CV 2025-02 conditional novelty 6.0 of 10

    FIGConv uses factorized 3D grids and global convolutions to predict car surface pressure and drag, reporting R2=0.957 on DrivAerNet drag and 0.89% pressure error on Ahmed body.

  28. Graph Transformers for inverse physics: reconstructing flows around arbitrary 2D airfoils

    cs.LG 2025-01 conditional novelty 6.0 of 10

    A hybrid message-passing and linear-attention Graph Transformer reconstructs 2D airfoil flow fields from surface pressure alone, achieving high test accuracy on a new open dataset.

  29. Point Cloud Neural Operator for Parametric PDEs on Complex and Variable Geometries

    math.NA 2025-01 conditional novelty 6.0 of 10

    A point cloud neural operator combining Fourier integral and least-squares gradient layers approximates PDE solution maps on variable geometries with reported test errors around 0.17 percent to 7 percent.

  30. Data-Driven Radio Propagation Modeling using Graph Neural Networks

    cs.LG 2025-01 conditional novelty 6.0 of 10

    A graph neural network with grid and ray-tracing edges, trained with masked outputs on real-world 4G measurements, generates radio coverage maps with 8.5 dB RMSE on outdoor data and 0.18 s GPU inference.

  31. Fusion-DeepONet: A Data-Efficient Neural Operator for Geometry-Dependent Hypersonic and Supersonic Flows

    cs.LG 2025-01 conditional novelty 6.0 of 10

    Fusion-DeepONet, a DeepONet variant that modulates every trunk layer with branch features, predicts geometry-dependent hypersonic and supersonic flows with fewer parameters and better accuracy on irregular grids than ...

  32. Incremental Hierarchical Tucker Decomposition

    math.NA 2024-12 conditional novelty 6.0 of 10

    The paper presents BHT-l2r, a batch hierarchical Tucker decomposition, and HT-RISE, its incremental update, with error bounds and experiments showing improved compression and generalization on PDE and image datasets.

  33. A Multi-Fidelity Graph U-Net Model for Accelerated Physics Simulations

    cs.LG 2024-12 conditional novelty 6.0 of 10

    A shared-parameter graph U-Net that couples coarse and fine mesh simulations during training predicts high-fidelity PDE solutions more accurately than single-fidelity GNNs or multi-fidelity transfer learning.

  34. A data-driven learned discretization approach in finite volume schemes for hyperbolic conservation laws and varying boundary conditions

    math.NA 2024-12 conditional novelty 6.0 of 10

    A learned finite-volume discretization, trained on fine-grid reference solutions, reproduces fine-grid accuracy for hyperbolic conservation laws on coarse meshes.

  35. Scaling Large-scale GNN Training to Thousands of Processors on CPU-based Supercomputers

    cs.DC 2024-11 conditional novelty 6.0 of 10

    SuperGCN trains full-batch GCNs on CPU supercomputers with minimum-vertex-cover hybrid aggregation and Int2 quantization, reaching 8,192 MPI ranks and beating prior CPU and GPU frameworks on several large datasets.

  36. Learning Pore-scale Multi-phase Flow from Experimental Data with Graph Neural Network

    physics.flu-dyn 2024-11 conditional novelty 6.0 of 10

    LSE-MGN, a graph neural network with long- and short-edge message passing, predicts pore-scale gas/liquid evolution from experimental micro-CT data with roughly 9 to 10 percent surface-area error over short autoregres...

  37. UBSoft: A Simulation Platform for Robotic Skill Learning in Unbounded Soft Environments

    cs.RO 2024-11 conditional novelty 6.0 of 10

    UBSoft is a simulation platform that uses robot-centered adaptive resolution to make unbounded soft-environment robot skill learning tractable, with benchmark tasks and sim-to-real demos.

  38. Data-free neural PDE solvers based on Graph Neural Networks and weak forms

    cs.CE 2026-07 conditional novelty 5.0 of 10

    A graph-neural-network PDE solver trained on the weak-form force residual — no simulation data — reports residual convergence below 1% on unseen load cases and one modified geometry, with residual-based test-time refinement.

  39. Recurrent U-Net-Based Graph Neural Network (RUGNN) for Accurate Deformation Predictions in Sheet Material Forming

    cs.LG 2025-07 conditional novelty 5.0 of 10

    RUGNN, a recurrent U-Net graph neural network with a node-to-surface contact feature, predicts sheet metal deformation fields across stamping timesteps with lower accumulated error than three GNN baselines on two FE-b...

  40. PDEfuncta: Spectrally-Aware Neural Representation for PDE Solution Modeling

    cs.LG 2025-06 conditional novelty 5.0 of 10

    A Fourier-based weight modulation for shared INR networks improves reconstruction of high-frequency PDE fields and enables bidirectional inference between paired solution spaces.

  41. SlotPi: Physics-informed Object-centric Reasoning Models

    cs.CV 2025-06 conditional novelty 5.0 of 10

    SlotPi combines a learned Hamiltonian energy module with spatiotemporal attention to improve object-centric video prediction and visual question answering on several datasets.

  42. Deep Learning Weather Models for Subregional Ocean Forecasting: A Case Study on the Canary Current Upwelling System

    physics.ao-ph 2025-05 conditional novelty 5.0 of 10

    An adapted GraphCast graph neural network trained on satellite sea surface temperature outperforms ConvLSTM and the GLORYS reanalysis for medium-range forecasts in the Canary Current upwelling system.

  43. Learning Effective Dynamics across Spatio-Temporal Scales of Complex Flows

    cs.LG 2025-02 conditional novelty 5.0 of 10

    A graph-neural-network autoencoder combined with a transformer predicts 2D cylinder and backward-facing-step flows on unstructured meshes, matching OpenFOAM fields while running about 100 to 900 times faster.

  44. DoMINO: A Decomposable Multi-scale Iterative Neural Operator for Modeling Large Scale Engineering Simulations

    cs.LG 2025-01 conditional novelty 5.0 of 10

    DoMINO is a point-cloud-based neural operator that predicts aerodynamic surface and volume fields on large automotive meshes with moderate accuracy and holds up on out-of-distribution car designs.

  45. A Graph Neural Network Surrogate Model for Multi-Objective Fluid-Acoustic Shape Optimization

    physics.flu-dyn 2024-12 reject novelty 5.0 of 10

    A GNN surrogate predicts airfoil flow fields and boundary layers accurately enough to drive a multi-objective noise-and-lift shape optimizer, though the headline 13.9% noise reduction is a comparison between Pareto po...

  46. Predicting Change, Not States: An Alternate Framework for Neural PDE Surrogates

    cs.LG 2024-12 conditional novelty 5.0 of 10

    Predicting the temporal derivative and integrating it with an ODE solver improves accuracy and stability of neural PDE surrogates compared with direct next-state prediction.

  47. Nonlinear Reduced-Order Modeling of Compressible Flow Fields Using Deep Learning and Manifold Learning

    physics.flu-dyn 2024-12 conditional novelty 5.0 of 10

    A deep-learning plus manifold-learning reduced-order model predicts transonic airfoil pressure fields and shock positions with about 3.5% error, outperforming POD and ISOMAP baselines when enough training data is available.

  48. AH-UGC: Adaptive and Heterogeneous-Universal Graph Coarsening

    cs.SI 2025-05 conditional novelty 4.0 of 10

    AH-UGC produces multiple coarsened graphs from one sorted hashed list and keeps heterogeneous node types pure by coarsening each type separately.

  49. PeSANet: Physics-encoded Spectral Attention Network for Simulating PDE-Governed Complex Systems

    cs.LG 2025-05 conditional novelty 4.0 of 10

    PeSANet reports lower RMSE and longer high-correlation times than FNO, F-FNO, FactFormer, and PeRCNN on 2D Burgers, FitzHugh-Nagumo, Gray-Scott, and Navier-Stokes benchmarks trained on 2 to 5 trajectories.

  50. Prototype-enhanced prediction in graph neural networks for climate applications

    cs.LG 2025-04 reject novelty 4.0 of 10

    Adding a target-derived prototype as an input improves GNN footprint emulator scores, but only under an oracle assignment that leaks the true label.

  51. Graph Neural Network for Cerebral Blood Flow Prediction With Clinical Datasets

    eess.IV 2024-11 conditional novelty 4.0 of 10

    A GNN adapted from cardiovascular modeling predicts Poiseuille-model blood flow and pressure in new brain vessel networks from MRA-derived graphs, reaching Pearson correlations of 0.727 (pressure) and 0.824 (flow).

  52. X-MeshGraphNet: Scalable Multi-Scale Graph Neural Networks for Physics Simulation

    cs.LG 2024-11 conditional novelty 4.0 of 10

    X-MeshGraphNet trains large physics-surrogate GNNs by partitioning graphs with halo regions, builds graphs from point clouds instead of meshes, and demonstrates scaling to 512 GPUs on car aerodynamics.

  53. PINN-FEM: A Hybrid Approach for Enforcing Dirichlet Boundary Conditions in Physics-Informed Neural Networks

    cs.LG 2025-01 reject novelty 3.0 of 10

    PINN-FEM enforces Dirichlet boundary conditions in PINNs by blending neural network fields with finite element shape functions in a boundary layer, but the 2D extension is ambiguous and the experimental comparisons ar...

Pith tools