- Rating
- 2
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 1
- Soundness
- 2
Read review
Paper summary
The paper proposes a topology-informed self-supervised model for predicting the mechanical properties of triply periodic minimal-surface metamaterials. A 3D denoising masked autoencoder with a ViT backbone is pretrained on signed-distance fields, after which its frozen representations are evaluated on mechanical-property tasks. Persistent homology modifies the model in two ways: patch-masking probabilities are weighted by the persistence of local , , and features, and an auxiliary head predicts the log feature count in each homology dimension from the pooled latent, with weight . The motivation is that small topological changes can produce much larger changes in effective stiffness, as illustrated by the cut/merge example in Figure 1. Experiments use 13,114 MetaDB structures and 18,549 interpolated TPMS structures, with homogenized stiffness tensors as labels. The learned representations are tested on family classification, classification, unsupervised clustering, and regression against nine baselines: MVAE, SVAE, VICReg, I-JEPA, MAE, DMAE, ColorMAE, DaM, and ScalableTopo. The clearest gains are in unsupervised clustering, where rises from 0.27 to 0.41 on Interpolation and from 0.15 to 0.32 on MetaDB. Family classification improves on MetaDB, while regression remains comparable to DMAE. The appendix includes component ablations, sensitivity sweeps, and a latent-alignment analysis.
Critical strengths
-
The motivation is clear, and the empirical results are strong. The teaser figure cut/merge example shows that a change of less than 1% in volume can alter the persistence diagram while changing stiffness by 7–10%. The experiments cover two datasets, nine baselines, four downstream tasks, a five-seed ablation with standard deviations, and a hyperparameter sweep. The paper also reports cases where other methods perform better, including the Interpolation linear probes and MetaDB regression.
-
The PH objective produces a clear clustering gain. For unsupervised clustering, increases from 0.272 to 0.412 on Interpolation and from 0.147 to 0.323 on MetaDB. These improvements are larger than the variation across seeds and are reflected in both AccMatch and macro-F1 in Table S3. Appendix A.4 also describes the persistent-homology pipeline in enough detail to make the experiment reproducible.
Critical weaknesses
- It is difficult to tell which part of the method is new, and the ablation does not support the paper’s explanation. The paper has no related-work section, so Table 1 is the only comparison. According to the table, PH-based supervision is already used by TopoAE (Moor et al., 2020), Hofer et al. (2019), and PH-SSL (Luo et al., 2023), however these methods are not discussed in the text. This seems to leave persistence-guided masking as the main new component. Table S5 does not show a measurable benefit from that component. On Interpolation, the PH head alone reaches , compared with for the full model. Masking alone reaches 0.251, below the 0.272 baseline. On MetaDB, the three results, 0.315, 0.315, and , are effectively indistinguishable. The PH head also reduces regression performance there, from without it to 0.899 with it. I therefore do not see evidence for the claimed “complementary roles” of the two components. The method notation also needs to be cleaned up. The total loss appear as , , and , with an additional introduced in Appendix A.10. The masking ratio is denoted by both and . No values are given for the dimension weights , and the persistence threshold , which determines the regression targets, appears only in Appendix A.4.
- The evaluation protocol is not clear enough to support the reported results. Appendix A.10 says that hyperparameters were selected using downstream-task metrics, but does not say whether those metrics came from the validation split. This matters because the results are quite sensitive to the chosen values: ranges from 0.097 to 0.461 across and from 0.150 to 0.317 across . The paper also does not explain how the nine baselines were tuned. There is also an unresolved discrepancy between tables. The selected configuration obtains in Table S6, but the corresponding value in Tables S3 and S5 are . These appear to be results for the same configuration and should either agree or be clearly distinguished. Since the PH head predicts three feature counts, Tasks 2–4 should include a probe using only This would show the source of the downstream gains. Finally, several figures make the comparisons harder to assess. Figure 5 connects four unrelated descriptors with lines, Figure 3 places ten methods on one radar chart, and Figure 4 groups different metrics from the two datasets under a shared label. Separate plots with explicit axes would make the results easier to verify.
Questions for authors
None
10 comments
Response 2/7: Baseline analysis, four result tables, and manuscript revisions
Reviewer #3 (VXCu) — Response 2/7
(Continue from Response 1/7)
These results support two main observations. First, our proposed mechanisms are more effective for mechanics-oriented representation learning in most evaluated metrics. For example, on MetaDB, TopoAE performs well in family classification (kNN macro-F1 = 0.978) but achieves a clustering of only 0.012, compared with 0.323 for our model. This may suggest that preserving topological relationships alone does not necessarily produce representations that capture mechanical-property variations.
Second, directly using PH descriptors does not provide consistently strong performance across different tasks. For example, on MetaDB, PH Descriptor achieves a regression of 0.423, compared with 0.899 for our model. Although PH Descriptor achieves a higher clustering on Interpolation TPMS (0.492 versus 0.412), our method performs better on the corresponding clustering accuracy and macro-F1, as well as the classification and regression tasks. These results, together with our ablation studies, suggest that integrating PH information into the learning process provides broader benefits than directly using PH descriptors alone.
Specifically, persistence-guided masking directs the encoder toward regions associated with persistent topological features, while the auxiliary PH head encourages the learned representation to retain global topological information. Together, these mechanisms incorporate PH into the learning of full 3D geometry representations, with the goal of capturing structural characteristics relevant to mechanical response.
The result tables (Task1-Task4) for comparison between the three baselines and our proposed Full model are shown below:
Task1 - Family Classification (new results were updated to Table 2 in the revised manuscrip)
| Model | MetaDB F Lin ↑ | MetaDB F kNN ↑ | Interpolation Linear Strict Acc ↑ | Interpolation Linear Relaxed Acc ↑ | Interpolation kNN Strict Acc ↑ | Interpolation kNN Relaxed Acc ↑ |
|---|---|---|---|---|---|---|
| TopoAE | 0.961 | 0.978 | 0.750 | 0.774 | 0.772 | 0.897 |
| PH Descriptor | 0.290 | 0.521 | 0.013 | 0.023 | 0.210 | 0.261 |
| PH-SSL | 0.988 | 0.974 | 0.872 | 0.881 | 0.830 | 0.939 |
| Ours | 0.994 | 0.985 | 0.867 | 0.880 | 0.831 | 0.939 |
Task 2 - Classification (new results were updated to Table S2 in the revised manuscript)
| Model | MetaDB kNN Bal. Acc ↑ | MetaDB kNN Macro-F1 ↑ | MetaDB Linear Bal. Acc ↑ | MetaDB Linear Macro-F1 ↑ | Interpolation kNN Bal. Acc ↑ | Interpolation kNN Macro-F1 ↑ | Interpolation Linear Bal. Acc ↑ | Interpolation Linear Macro-F1 ↑ |
|---|---|---|---|---|---|---|---|---|
| ScalableTopo | 0.954 | 0.954 | 0.960 | 0.957 | 0.856 | 0.856 | 0.916 | 0.916 |
| TopoAE | 0.950 | 0.950 | 0.930 | 0.930 | 0.825 | 0.825 | 0.877 | 0.877 |
| PH Descriptor | 0.666 | 0.665 | 0.443 | 0.414 | 0.839 | 0.840 | 0.843 | 0.844 |
| PH-SSL | 0.936 | 0.936 | 0.942 | 0.942 | 0.871 | 0.871 | 0.912 | 0.912 |
| Ours | 0.951 | 0.951 | 0.964 | 0.964 | 0.887 | 0.887 | 0.924 | 0.925 |
Task 3 - Clustering (new results were updated to Table S3 in the revised manuscript)
| Model | MetaDB ↑ | MetaDB AccMatch ↑ | MetaDB Macro-F1 ↑ | Interpolation ↑ | Interpolation AccMatch ↑ | Interpolation Macro-F1 ↑ |
|---|---|---|---|---|---|---|
| ScalableTopo | 0.199 | 0.543 | 0.533 | 0.158 | 0.422 | 0.395 |
| TopoAE | 0.012 | 0.471 | 0.464 | 0.188 | 0.476 | 0.477 |
| PH Descriptor | 0.026 | 0.381 | 0.306 | 0.492 | 0.554 | 0.513 |
| PH-SSL | 0.023 | 0.593 | 0.586 | 0.286 | 0.515 | 0.502 |
| Ours | 0.323 | 0.660 | 0.673 | 0.412 | 0.582 | 0.576 |
Task 4 - Regression (new results were updated to Table S4 in the revised manuscript)
| Model | MetaDB MAE ↓ | MetaDB RMSE ↓ | MetaDB ↑ | Interpolation MAE ↓ | Interpolation RMSE ↓ | Interpolation ↑ |
|---|---|---|---|---|---|---|
| ScalableTopo | 0.008 | 0.041 | 0.932 | 0.003 | 0.005 | 0.992 |
| TopoAE | 0.042 | 0.092 | 0.685 | 0.004 | 0.006 | 0.987 |
| PH Descriptor | 0.046 | 0.124 | 0.423 | 0.010 | 0.014 | 0.934 |
| PH-SSL | 0.010 | 0.047 | 0.917 | 0.003 | 0.005 | 0.992 |
| Ours | 0.009 | 0.050 | 0.899 | 0.003 | 0.005 | 0.992 |
Corrections:
We have added the following paragraph to Section 3.2 (Baselines) of the revised manuscript:
(lines 328-337) “Additionally, we include three topology-informed baselines: ScalableTopo, which uses scalable persistent-homology regularization; TopoAE, which preserves topological information in learned representations; and PH-SSL, which incorporates persistent-homology-based self-supervision to learn topology-aware representations. We also include PH-based descriptors (, , and persistence counts, together with volume fraction), which directly use persistent-homology feature counts as structural descriptors.”
The corresponding baseline results have also been added to Table 2 and Tables S2–S4 in the revised manuscript.
Response 3/7: Roles of the two Mechanisms and Ablation Interpretation
Reviewer #3 (VXCu) — Response 3/7
Reviewer comment:
Table S5 does not show a measurable benefit from that component. On Interpolation, the PH head alone reaches , compared with for the full model. Masking alone reaches 0.251, below the 0.272 baseline. On MetaDB, the three results, 0.315, 0.315, and , are effectively indistinguishable. The PH head also reduces regression performance there, from without it to 0.899 with it. I therefore do not see evidence for the claimed “complementary roles” of the two components.
Response: Thank you for highlighting the benefit for each of our proposed component. We agree that, on a given metric, the full model is often only marginally better than the stronger single component, and that the PH head reduces regression on MetaDB (0.937 → 0.899). We have therefore revised the term “complementary roles” to describe more precisely what Table S5 shows: the two mechanisms contribute in a dataset-dependent way, and their combination removes the need to choose one per dataset.
On Interpolation TPMS, the gain comes mainly from the PH head. Adding the head to the baseline raises clustering from 0.272 to 0.405, whereas masking alone slightly lowers it (0.251). On MetaDB, masking provides benefits that the head does not. Masking alone raises from 0.147 to 0.315, gives the best family kNN macro-F1 (0.993), and gives the best regression (0.937), compared with 0.907 for the head-only model. Thus, the most effective single component differs between the two datasets: head-only ranks second on Interpolation TPMS (average rank 2.06) but third on MetaDB (2.63), while masking-only ranks last on Interpolation TPMS (3.38) but second on MetaDB (2.00).
The full model does not dominate every entry, but it is the most consistent across both datasets: it achieves the best mean on 12 of the 16 metric–dataset entries and the best average rank on both datasets (1.38). We have revised the text to present this as a dataset-dependent tradeoff rather than as uniform complementarity.
Corrections:
(Lines 427-428) In Section 3.4: we revised the conclusion of ablation study to “their advantages are dataset-dependent”
(Lines 477-483) In Section 4: Ablation studies show that the contributions of the two topology-aware components are dataset-dependent: the PH head contributes more on Interpolation TPMS, whereas persistence-guided masking contributes more on MetaDB. Combining both components removes the need to select one per dataset, and the full model achieves the most consistent performance across both datasets.
(Lines 533-539) “Nevertheless, our study has two main limitations that need further investigations. First, the persistent-homology targets are computed on the finite voxel domain of each sample without periodic identification, so topological features that close only across the periodic boundary are not counted; a periodic filtration would align more directly with the periodic mechanical connectivity of TPMS structures. Second, physically meaningful descriptors such as volume fraction strongly determine axial stiffness, and how to combine such descriptors with learned topology-aware representations remains an open direction for future work.”
Reviewer comment:
The method notation also needs to be cleaned up. The total loss appears as , , and , with an additional introduced in Appendix A.10.
Response: We apologize for the confusion caused by the inconsistent notation. We have standardized the notation for the total training objective throughout the manuscript as:
where denotes the reconstruction loss, represents the persistent-homology supervision loss, and controls its contribution.
Corrections:
(Lines 179-181) Figure 2 (Page 4, Line 180): Changed "" to "".
(Lines 179-181) Appendix A.3 (Page 19, Line 995): Changed "" to "".
(Lines 179-181) Appendix A.10 (Page 28, Line 1477): "the PH-loss weight , which are most relevant ".
These corrections ensure consistent notation across the manuscript without changing the actual training objective or experimental implementation. We thank the reviewer for pointing out these inconsistencies.
Response 1/7: Novelty, related work, and introduction of additional baselines
Reviewer #3 (VXCu) — Response 1/7
We sincerely thank the reviewer for the time and effort spent on our manuscript and for the constructive feedback, which has improved it substantially. The comments on novelty led us to add a related-work discussion and a TopoAE baseline, which clarify how our approach differs from existing PH-based methods. The suggested probe on PH counts and volume fraction, together with the closer reading of the ablation, helped us describe the roles of the two mechanisms more accurately: their contributions are dataset-dependent, and the full model is the most consistent across both datasets. We have also clarified the evaluation protocol, unified the notation, and redrawn Figures 3–5 as suggested. We address each point below.
Reviewer comment:
Response: We thank the reviewer for pointing out these relevant works. In the original manuscript, we included the related-work discussion within the Introduction. To make the comparison with prior studies clearer, we have added a separate related-work section discussing these methods in the revised manuscript. We have also expanded it to include TopoAE, Hofer et al., and PH-SSL, and to clarify the differences in motivation and methodology.
For motivation, TopoAE (Moor et al., 2020) aims to preserve topological relationships between the input data and their latent representations; Hofer et al. (2019) focuses on controlling latent-space connectivity for one-class learning; PH-SSL introduces persistent topological fingerprints to capture more fine-grained relationships between molecules.Our work is motivated by the relationship between structural connectivity and mechanical response in 3D TPMS metamaterials. Local changes in connectivity can alter load-transfer pathways and effective stiffness, even when the overall geometric variation is small. As illustrated by the controlled Cut/Merge experiments in Fig. 1 and Appendix Sec. A.1, localized structural modifications produce noticeable changes in both persistent-topological descriptors and directional mechanical responses.
And for methodology, these approaches incorporate PH into representation learning in different ways. TopoAE uses a topological loss to preserve relationships between the input and latent spaces; Hofer et al. (2019) introduces PH-based regularization to control latent-space connectivity; PH-SSL uses persistent topological fingerprints to provide self-supervision through topology prediction and contrastive learning. Our method instead incorporates PH into two stages: First, persistence-guided masking assigns higher masking probabilities to patches associated with persistent topological features, encouraging the encoder to learn their structural context. Second, an auxiliary PH prediction head supervises the latent representation using the counts of persistent features in H0, H1, and H2. Together, the two mechanisms encourage learning representations to capture structural connectivity relevant to mechanical response.
We’ve updated the explanations for the topology informed methods in Related Work:
(lines 091-099) “Prior topology-aware methods, as summarized in Table 1, have largely incorporated PH to encode structural connectivity in 2D (Hu et al., 2023; Wang et al., 2020a) or in generic geometric settings, through topological pooling and representations (Chen & Gel, 2023; Tola et al., 2025). Several works further use PH directly as a training signal: TopoAE (Moor et al., 2020) and Hofer et al. (2019) preserve the topology of data or latent manifolds, and PH-SSL (Luo et al., 2023) uses persistence-based fingerprints for molecular self-supervision. These methods, however, are not designed around mechanical response.”
To demonstrate that these mechanisms are well suited to this objective, we added TopoAE and PH-SSL as baseline comparisons to test whether their topology preservation methods are sufficient, and we also considered a direct PH-descriptor (log(H0), log(H1), log(H2), volume fraction) baseline to test whether simple geometric features alone are sufficient. Overall, our method achieves stronger performance than the three topology-informed baselines across most mechanics-related metrics on both datasets.
(Continue in Response 2/7)