- Rating
- 1
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
The paper introduces an alternative method to construct a hierarchical graph for graph-based neural simulators. Here, the authors use mechanical information from the dynamics, such as stiffness, loads, and stress, to reweight the mesh edges. They then merge neighboring node groups based on changes in structural entropy. The resulting tree then defines the hierarchy used by the learned simulator.
Critical strengths
- The author showed that the structural-entropy approach performs somewhat better than the random coarsening and the position attention on the BilateralStamping task.
Critical weaknesses
-
As stated in the discussion section, the evaluation is still limited to two solid-mechanics stamping tasks. Whether the approach can be extended to other dynamics remains unclear. In the paper's own experiment, on the larger UnilateralStamping task, it does not show a clear gain over the baselines.
-
More importantly, the paper does not explain how the coefficients in the mechanical weighting were chosen. This leads to a critical concern: how to select the "mechanical information" and its corresponding coefficients for different dynamics, e.g., fluid dynamics, which would severely limit the method’s broader applicability.
-
The hierarchy comparison only includes random coarsening and position attention. I would like to see comparisons with M4GN [1], which uses modal-decomposition features for grouping, and AMG-based approaches such as ROBIN [2]. Without these comparisons, it is unclear whether structural entropy is more useful than existing approaches or simply performs similarly.
-
The figures need substantial revision. For example, in Figure 5, the colors do not clearly highlight the groups, and it is unclear what the authors want to show or how the figure supports their approach.
[1] B. Lei, et al. M4GN: Mesh-based Multi-segment Hierarchical Graph Network for Dynamic Simulations. TMLR, 2025.
[2] T. Würth, et al. Diffusion-Based Hierarchical Graph Neural Networks for Simulating Nonlinear Solid Mechanics. NeurIPS, 2025.
Questions for authors
See weaknesses
3 comments
Response to Reviewer GjqM (Part 2 of 2)
W.4 Comparison with M4GN and AMG
We have added M4GN, which uses modal features for grouping, and ROBIN, which uses AMG, to Table 1. Results on both stamping tasks follow.
BilateralStamping relative L₂ errors in ×10⁻² units. M denotes metal; R denotes rubber. Lower is better.
| Method | Geo M | Geo R | Stress M | Stress R | PEEQ M | PEEQ R |
|---|---|---|---|---|---|---|
| Geo FNO | 1.25 | 1.17 | 3.90 | 16.12 | 42.69 | 47.31 |
| GINO | 2.99 | 2.43 | 5.43 | 26.21 | 55.23 | 64.09 |
| GNO | 2.85 | 2.81 | 4.63 | 25.24 | 58.30 | 59.09 |
| LSM | 5.98 | 6.15 | 5.56 | 36.16 | 70.01 | 73.47 |
| Galerkin | 5.99 | 5.94 | 5.63 | 34.42 | 69.62 | 73.35 |
| FactFormer | 2.97 | 1.05 | 4.59 | 13.78 | 62.66 | 42.18 |
| OFormer | 3.34 | 2.68 | 4.95 | 31.78 | 55.51 | 51.57 |
| ONO | 3.01 | 2.75 | 4.64 | 32.91 | 51.76 | 52.31 |
| Transolver | 0.83 | 0.81 | 3.46 | 14.25 | 35.92 | 32.76 |
| GraphSAGE | 2.27 | 2.29 | 4.40 | 21.96 | 49.36 | 44.95 |
| MGN | 1.95 | 1.97 | 4.21 | 20.59 | 46.46 | 41.41 |
| Unisoma | 0.57 | 0.52 | 2.78 | 10.39 | 28.17 | 22.65 |
| M4GN | 0.64 | 0.63 | 2.70 | 11.27 | 27.44 | 24.82 |
| ROBIN | 1.13 | 2.79 | 3.59 | 27.87 | 38.57 | 56.59 |
| EntroMesh | 0.50 | 0.51 | 2.57 | 10.18 | 26.14 | 22.71 |
UnilateralStamping relative L₂ errors in ×10⁻² units. Lower is better.
| Method | Geo M | Geo R | Stress M | Stress R | PEEQ M | PEEQ R |
|---|---|---|---|---|---|---|
| Geo FNO | 1.09 | 1.54 | 11.13 | 11.24 | 37.71 | 35.74 |
| GINO | 1.88 | 2.19 | 12.62 | 14.38 | 46.51 | 48.85 |
| GNO | 4.26 | 4.24 | 13.51 | 18.14 | 64.13 | 68.59 |
| LSM | 3.67 | 3.57 | 13.27 | 13.35 | 58.89 | 60.26 |
| Galerkin | 10.07 | 10.10 | 17.25 | 27.94 | 79.86 | 85.55 |
| FactFormer | 1.78 | 1.66 | 10.95 | 10.35 | 38.86 | 30.22 |
| OFormer | 1.19 | 0.98 | 11.03 | 9.81 | 32.69 | 26.75 |
| ONO | 4.33 | 4.37 | 13.77 | 13.11 | 51.40 | 47.48 |
| Transolver | 0.79 | 0.75 | 8.39 | 8.62 | 29.38 | 26.93 |
| GraphSAGE | 2.75 | 2.60 | 10.59 | 11.89 | 54.41 | 51.06 |
| MGN | 3.91 | 3.83 | 12.52 | 14.17 | 58.05 | 56.94 |
| Unisoma | 0.71 | 0.63 | 8.07 | 7.73 | 25.91 | 21.14 |
| M4GN | 0.77 | 0.66 | 7.87 | 7.52 | 23.19 | 19.62 |
| ROBIN | 4.39 | 5.19 | 14.28 | 20.05 | 73.53 | 80.37 |
| EntroMesh | 0.57 | 0.51 | 7.56 | 6.80 | 20.75 | 17.47 |
EntroMesh outperforms M4GN and ROBIN across all six fields on both tasks. Table 1 compares complete models, while Table 2 compares hierarchy rules under the same architecture and node budgets.
W.5 Figure Clarity
We have redesigned the original Figure 5, now Figure 7, and added analysis in Section 4.3. It follows one real BilateralStamping input through physical inputs, edge weighting, and coarse modules.
Thank you for your review
We have uploaded the revised paper and the corresponding responses. The revised parts of the paper are marked in red. We hope our responses can address some of your concerns. We are also very grateful for your valuable feedback on our paper. If our response can satisfy you, we would be extremely honored that you could revise your previous opinion on our paper.
Response to Reviewer GjqM (Part 1 of 2)
W.1 Generalization Beyond Stamping
We have added CylinderFlow experiments on autoregressive velocity and pressure prediction over unstructured meshes. Results are included in Table 1, and Figure 10 in Appendix C.2 illustrates the corresponding physical inputs and hierarchy.
CylinderFlow errors, reported as mean ± standard deviation. Lower is better.
EntroMesh has the lowest mean error at all three horizons, reducing complete trajectory error by 11.76% relative to M4GN. This provides evidence of applicability beyond solid stamping.
W.2 Effectiveness on UnilateralStamping
We have added M4GN and ROBIN comparisons in Table 1 and analyzed the field results in Section 4.2. The main comparisons are summarized below; Table 1 contains all baselines.
UnilateralStamping relative L₂ errors in ×10⁻² units. M denotes metal; R denotes rubber. Lower is better.
EntroMesh achieves state of the art performance among all methods in Table 1, with the lowest error across all six fields. Aggregate error decreases by 16.42% relative to Unisoma, covering improvements in geometry, stress, and plastic response.
W.3 Weight Selection and Cue Ablation
We have clarified cue selection in Section 3.2 and Figure 3 and added weighting ablations in Table 5 of Appendix F.1. Loads and boundaries describe operating conditions, stiffness represents material coupling, energy identifies concentrated response, and stress jumps identify candidate module boundaries. Available cues are normalized into dimensionless graph weights.
Both stamping tasks use stiffness, load, and energy coefficients of 1.0, boundary coefficient 0.75, and stress consistency coefficient 0.5. These empirical settings provide reference amplification, moderate boundary amplification, and smooth attenuation across stress jumps. Figure 6 reports local sensitivity.
BilateralStamping weighting ablations. Relative L₂ errors are in ×10⁻² units. M denotes metal; R denotes rubber. Lower is better.
The full model achieves the lowest error across all fields. Removing load and boundary weighting increases rubber PEEQ error by 6.16% and 6.91%; geometry weights alone increase it by 7.53%. Rubber stress also worsens. These comparisons establish the individual contributions of load and boundary weighting and the combined benefit over geometry alone.