- Rating
- 3(+1)
- Confidence
- 3
- Informativeness
- 2(+1)
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
This paper presents HybridOSS, a hybrid framework for fine-resolution PDE prediction using a coarse-grid numerical solver. The method decomposes the state into orthogonal resolved and unresolved spectral components. A differentiable numerical solver advances the resolved state on a coarse grid, while a learned closure module corrects the resolved dynamics and a refinement module updates the unresolved state. Spectral projections constrain the two learned corrections to complementary subspaces while allowing cross-scale interactions. The method is evaluated on Kuramoto–Sivashinsky, Allen–Cahn, and 3D Navier–Stokes equations. The reported results show improved prediction accuracy over the evaluated baselines, along with computational benefits in larger-scale simulations. Ablation studies further examine the contributions of the closure and refinement modules, spectral projection, and different coarsening factors.
Critical strengths
The motivation for decoupling the numerical solver resolution from the predicted state resolution is reasonable and well justified, providing a practical way to balance computational efficiency and fine-scale prediction accuracy. The paper also presents a series of experiments and ablation studies that support the effectiveness and feasibility of the proposed design. I consider these to be the main strengths of this work.
Critical weaknesses
- My main concern is the limited novelty of the proposed method. The first two contributions, including orthogonal scale decomposition, separate modeling of resolved and unresolved dynamics, and combining coarse-grid solvers with neural corrections, have been explored in prior work. The proposed framework appears to be more of a combination of existing techniques. I believe the authors should clarify what new technical contributions this combination offers compared with existing methods. What I mean is that the authors should more clearly explain why they chose this particular combination of components. I believe this is where the paper's novelty could be most clearly demonstrated.
Relevant references (including some already cited in the manuscript):
Marion & Temam (1989), Nonlinear Galerkin Methods. https://doi.org/10.1137/0726063
Devulder & Marion (1992), A Class of Numerical Algorithms for Large Time Integration: The Nonlinear Galerkin Methods. https://doi.org/10.1137/0729028
Hughes et al. (1998), The Variational Multiscale Method—A Paradigm for Computational Mechanics. https://doi.org/10.1016/S0045-7825(98)00079-6
Margenberg et al. (2022), A Neural Network Multigrid Solver for the Navier–Stokes Equations. https://arxiv.org/abs/2008.11520
Lai et al. (2024), Neural Downscaling for Complex Systems: From Large-scale to Small-scale by Neural Operator. https://arxiv.org/abs/2403.13016
- Following up on the previous point, the authors should use ablation studies to demonstrate why these components need to be combined in this particular way. The authors provide ablation studies for the Closure module, Refinement module, and orthogonal projection, but the role of cross-scale coupling is not directly examined. For example, removing the high-frequency input to the Closure module or the low-frequency input to the Refinement module would help determine whether these interactions are necessary. In addition, the current component ablations mainly focus on KS and AC. Extending these experiments to the 3D Navier–Stokes task would provide stronger evidence for why the proposed combination is effective.
Questions for authors
Please refer to the Critical Weaknesses section.
7 comments
W2: Validating Cross-Scale Coupling and Component Synergy through Extended Ablations
1. Cross-scale input ablations.
We added three ablations: masking the high-frequency input to Closure, masking the low-frequency input to Refinement, and masking both directions. Masking is applied before feature construction, so derived features are computed without direct access to the removed scale component. The cross-scale-input variants retain the same network architecture and training protocol. All results are reported as mean ± standard deviation across three independent training runs.
2. Results on KS and AC.
All three ablations increase the mean global error on both KS and AC. On AC, they also increase the mean resolved- and unresolved-scale errors. On KS, some individual scale-specific errors improve, but the global error still increases. These results support the benefit of cross-scale conditioning for full-state prediction.
| Configuration | KS Global | KS Resolved | KS Unresolved | AC Global | AC Resolved | AC Unresolved |
|---|---|---|---|---|---|---|
| w/o HF input to Closure | 0.1028±0.0030 | 0.0084±0.0006 | 0.9414±0.0475 | 0.0223±0.0065 | 0.0008±0.0006 | 0.2120±0.0118 |
| w/o LF input to Refinement | 0.1047±0.0024 | 0.0147±0.0007 | 0.3509±0.0462 | 0.0202±0.0019 | 0.0006±0.0002 | 1.0095±0.0033 |
| w/o cross-scale inputs | 0.1205±0.0029 | 0.0158±0.0007 | 0.7173±0.0214 | 0.0223±0.0033 | 0.0007±0.0004 | 1.0002±0.0002 |
| HybridOSS(full) | 0.0858±0.0022 | 0.0097±0.0003 | 0.4756±0.1151 | 0.0173±0.0010 | 0.0004±2.95e-5 | 0.1709±0.0510 |
3. Extension to three-dimensional NS.
We further extended the component and cross-scale-input ablations to 3D NS. All variants were trained only at and evaluated on the same test trajectories using the mean relative error over 125-step autoregressive rollouts.
The complete model achieves the lowest mean error at all four test viscosities. At , its error is , compared with , , and when masking the Closure high-frequency input, the Refinement low-frequency input, and both directions, respectively. Full-state and the projection ablation yield errors of and . Removing either Closure or Refinement also increases the error. At the unseen viscosities , the complete model reduces the mean error relative to Full-state by approximately , , and , respectively, and by – relative to the projection ablation. Masking either direction of cross-scale input also increases the mean error at each unseen viscosity, indicating that the benefits extend beyond the training viscosity.
| Configuration | ν=5×10⁻⁶ | ν=1×10⁻⁵ | ν=5×10⁻⁵ | ν=1×10⁻⁴ |
|---|---|---|---|---|
| w/o Closure | 0.8457 ± 0.0271 | 0.8450 ± 0.0276 | 0.8637 ± 0.0276 | 0.9168 ± 0.0261 |
| w/o Refinement | 0.4308 ± 0.0923 | 0.4482 ± 0.0944 | 0.5583 ± 0.1065 | 0.6919 ± 0.1183 |
| w/o projection | 0.3636 ± 0.0160 | 0.3465 ± 0.0160 | 0.4856 ± 0.0100 | 0.6566 ± 0.0127 |
| w/o HF input to Closure | 0.5050 ± 0.0648 | 0.4884 ± 0.0716 | 0.5396 ± 0.0568 | 0.6494 ± 0.0386 |
| w/o LF input to Refinement | 0.8290 ± 0.0448 | 0.8115 ± 0.0472 | 0.8334 ± 0.0347 | 0.8876 ± 0.0280 |
| w/o cross-scale inputs | 0.8176 ± 0.0483 | 0.8024 ± 0.0526 | 0.8164 ± 0.0426 | 0.8651 ± 0.0307 |
| Full state | 0.5571 ± 0.0177 | 0.4727 ± 0.0124 | 1.1414 ± 0.0170 | 1.5580 ± 0.0247 |
| HybridOSS(full) | 0.3151 ± 0.0337 | 0.2990 ± 0.0350 | 0.4157 ± 0.0169 | 0.5739 ± 0.0129 |
These results extend the evidence to three dimensions and support the complementary roles of branch separation, orthogonal projection, and cross-scale conditioning in the tested configurations. We will include these ablations and their variability across runs in the revised manuscript.
Thank you for your detailed responses and additional experiments. Most of my previous questions have been answered, and I now have a clearer and more positive view of the paper. However, I still have one important concern regarding the novelty of the method. I understand the differences between HybridOSS and previous methods, but I am still not sure why combining explicit unresolved-state evolution, a separate Closure module, and complementary orthogonal projections represents a distinct technical contribution. To me, this seems more like an effective combination of existing techniques rather than a new scientific insight. For example, I understand that the main motivation is to reduce computational cost while preserving cross-scale coupling. The first point is clear, and the paper provides convincing experimental evidence. However, for the second point, although the new ablations show that cross-scale inputs improve overall prediction accuracy, I am still unsure whether this means the model captures the underlying physical coupling between scales more accurately. The paper does not seem to provide a direct metric or analysis of these physical interactions. This is my main remaining concern, and I would be very willing to raise my score to 3 if the authors could clarify this point. I also noticed an interesting result in the KS ablations. Removing the high-frequency input to Closure improves the resolved error, while removing the low-frequency input to Refinement improves the unresolved error. Could this be related to the chaotic nature of KS, the coarsening factor, or error propagation between scales? I think this could be a useful case for further analysis, as it may help us better understand when cross-scale coupling is beneficial and when it becomes more difficult to learn.
We thank the reviewer for highlighting these connections.
W1: Clarifying the Novelty and Rationale of the Proposed Component Integration
1. Connection to prior work and our focus. We acknowledge that nonlinear Galerkin methods, their subsequent developments, and the variational multiscale method establish foundations for scale decomposition and cross-scale modeling. Our focus is maintaining the long-term evolution of a complete fine-grid state using a coarse-grid differentiable solver.
2. Distinction from related neural methods. DNN-MG feeds corrected fine-grid velocity information back into coarse-grid advancement through a restricted right-hand side, while Neural Downscaling maps large-scale temporal histories to small-scale states. HybridOSS explicitly propagates the unresolved state, learns its feedback through a separate Closure branch, and jointly trains complementary projected updates with a differentiable solver in the loop.
3. Why these components are combined. The design aims to reduce numerical cost while retaining coupled evolution across scales. Coarsening creates two tasks: correcting an existing numerical prediction on resolved scales and learning the evolution of unresolved scales without direct numerical advancement. Although one network can learn both tasks, two branches provide explicit responsibilities. Complementary projections restrict the frequency bands each branch can modify, while cross-scale inputs preserve information exchange between them. The Full-state control, projection ablation, and newly added cross-scale-input ablation provide complementary experimental support for this mechanism.
To the best of our knowledge, our work is the first to combine persistent unresolved-state evolution, cross-scale closure feedback, and complementary orthogonal projection constraints in a two-branch mechanism within a hybrid differentiable solver-in-the-loop framework, enabling coupled fine-grid evolution under coarse-grid numerical stepping.