- Rating
- 2
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 3
- Soundness
- 1
Read review
Paper summary
The paper introduces TopoForce, a PH-free, graph-based dimensionality-reduction method for learning representations that reveal multiscale topology. The central motivation is that many topology-preserving DR methods use persistent homology (PH) of the observed data as a target. When observed geometry is distorted by noise, nuisance variation, or nonuniform sampling, that target PH can be misleading, so directly preserving it may propagate distortion into the embedding. TopoForce instead builds a neighborhood graph, performs multiscale diffusion, and infers a pairwise topology scale for each pair of points. This scale reflects which diffusion modes most strongly distinguish the pair and how slowly those modes decay.
Critical strengths
I appreciate the move away from PH method, as it has known scalability and performance under noise issues. Experimental setup seems robust, especially the fact that it tests many scenarios with noise and distortion/variation.
Critical weaknesses
Scalability of TopoForce is still a major issue compared to other methods (table 3). It is also unclear how the parameters in Appendix C were chosen, and this may be a reason why TopoForce outperforms other methods? Finally, hhy experiments were not run on multiple seeds escapes me.
Questions for authors
- How were hyperparameters for TopoForce chosen? Is the parameter search conducted fair versus all other baseline? More information on this section is crucial to assess TopoForce fairly.
- Why was only one seed used for experiments? Is it because the original data is deterministic? What about noise injection?
- Is the method easily extendable to higher dimensional data?
4 comments
Response to Reviewer PLgL Q1
Q1. Hyperparameter selection and baseline fairness
"How were hyperparameters for TopoForce chosen? Is the parameter search conducted fair versus all other baseline? More information on this section is crucial to assess TopoForce fairly."
We have clarified the parameter-selection procedure and provide the effective configurations used in the experiments. In TopoForce, the neighbourhood size () and affinity bandwidth () are held fixed across the reported experiments. The diffusion reference time is selected adaptively from the knee of the Von Neumann entropy curve. The force coefficients (, , and ) are fixed, while the schedule is determined adaptively from the distribution of pairwise topology scales. Thus, the main force-design parameters are not independently tuned for individual benchmark conditions.
To further assess baseline sensitivity, we completed 256 parameter configurations across UMAP, t-SNE, Isomap, PHATE and Diffusion Maps on distorted and nonuniform bracelet and three-loop inputs. The scans cover neighbourhood sizes of 5, 15, 30 and 45, UMAP minimum distances from 0 to 0.9, t-SNE perplexities from 5 to 45, and diffusion times from 0 to 100, using seed 0. The resulting embeddings are available for visual comparison in the baseline hyperparameter sweeps (Rebuttal Figure R2 hyperlink).
These additional scans provide a broader assessment of how baseline representations respond to their respective parameter choices, rather than relying solely on default configurations.
Response to Reviewer PLgL Q2
Q2. Random seeds and noise injection
"Why was only one seed used for experiments? Is it because the original data is deterministic? What about noise injection?"
We acknowledge that reporting only one seed in the original submission did not allow the variability arising from random initialization and optimisation to be assessed. We have therefore expanded the evaluation to five algorithmic seeds (0–4), covering 20 synthetic conditions, four real-data conditions and 30 nuisance conditions, with 3,285 valid results across 13 methods. We report individual seed values, means and sample standard deviations.
On distorted three-loop data, TopoForce obtains a TRS of , compared with for DIPOLE, the strongest competing method by mean TRS. On nonuniform three-loop data, the corresponding scores are and . These results demonstrate consistent performance across algorithmic seeds, particularly on the distorted three-loop benchmark.
Updated results for Bracelet and Three loops
Topology recovery score (TRS, Eq. 9) under uniform (U), distorted uniform (UD), and nonuniform (NU) sampling. Each cell gives the mean ± sample SD over optimizer seeds 0–4 on the same input. Bold and italic values indicate the highest and second-highest distinct means, respectively, across all 13 methods before display rounding.
| Method | Bracelet U () | Bracelet UD () | Bracelet NU () | Three loops U () | Three loops UD () | Three loops NU () |
|---|---|---|---|---|---|---|
| TopoForce | 0.970 ± 0.000 | 0.877 ± 0.000 | 0.891 ± 0.001 | 0.975 ± 0.002 | 0.963 ± 0.002 | 0.978 ± 0.002 |
| UMAP | 0.944 ± 0.005 | 0.836 ± 0.028 | 0.619 ± 0.107 | 0.600 ± 0.141 | 0.668 ± 0.106 | 0.609 ± 0.219 |
| t-SNE | 0.947 ± 0.000 | 0.445 ± 0.000 | 0.465 ± 0.000 | 0.548 ± 0.000 | 0.454 ± 0.000 | 0.413 ± 0.000 |
| FLE | 0.509 ± 0.123 | 0.455 ± 0.145 | 0.439 ± 0.101 | 0.510 ± 0.144 | 0.461 ± 0.094 | 0.623 ± 0.131 |
| PHATE | 0.754 ± 0.048 | 0.567 ± 0.080 | 0.551 ± 0.119 | 0.504 ± 0.033 | 0.463 ± 0.034 | 0.474 ± 0.055 |
| Isomap | 0.919 ± 0.000 | 0.629 ± 0.000 | 0.851 ± 0.000 | 0.923 ± 0.000 | 0.329 ± 0.000 | 0.936 ± 0.000 |
| DIPOLE | 0.626 ± 0.063 | 0.643 ± 0.033 | 0.730 ± 0.017 | 0.947 ± 0.015 | 0.796 ± 0.105 | 0.957 ± 0.006 |
| TopoAE++ | 0.868 ± 0.035 | 0.600 ± 0.159 | 0.738 ± 0.230 | 0.845 ± 0.107 | 0.658 ± 0.133 | 0.903 ± 0.027 |
| RTD-AE | 0.780 ± 0.214 | 0.476 ± 0.209 | 0.889 ± 0.002 | 0.911 ± 0.001 | 0.732 ± 0.049 | 0.830 ± 0.134 |
| Diffusion Maps | 0.769 ± 0.000 | 0.346 ± 0.000 | 0.466 ± 0.000 | 0.000 ± 0.000 | 0.000 ± 0.000 | 0.323 ± 0.000 |
| Laplacian Eigenmaps | 0.650 ± 0.000 | 0.329 ± 0.000 | 0.436 ± 0.000 | 0.032 ± 0.000 | 0.000 ± 0.000 | 0.327 ± 0.000 |
| Stress majorization | 0.916 ± 0.002 | 0.792 ± 0.005 | 0.734 ± 0.010 | 0.908 ± 0.002 | 0.783 ± 0.001 | 0.903 ± 0.006 |
| tsNET | 0.307 ± 0.043 | 0.265 ± 0.069 | 0.208 ± 0.100 | 0.440 ± 0.143 | 0.389 ± 0.210 | 0.407 ± 0.085 |
To isolate algorithmic variability from variability introduced during data generation, the five-seed repetitions reuse identical sampled point coordinates and nuisance-noise realisations within each condition. Thus, the reported standard deviations quantify variability arising from initialization and optimisation conditional on the same input observations, rather than variability across independently generated noise realisations. Input identities and effective run configurations are recorded with the results in Rebuttal Figure R3 hyperlink, which also provides the five-seed results.
Response to Reviewer PLgL Q3
Q3. Applicability to higher-dimensional data
"Is the method easily extendable to higher-dimensional data?"
TopoForce operates on a neighbourhood graph, and its diffusion and pairwise-scale calculations are therefore not intrinsically restricted to low-dimensional input vectors. We provide empirical evidence through our high-dimensional nuisance experiments, which extend three synthetic structures with up to 2,048 additional Gaussian dimensions. Five algorithmic seeds are evaluated on each fixed input, allowing us to assess whether the recovered structures remain detectable as the input dimensionality increases.
For the output representation, the pairwise force formulation is mathematically applicable to embedding spaces of arbitrary dimension, since the pairwise interactions are defined through Euclidean distances and direction vectors in the embedding space. However, our current implementation and experiments focus on two- and three-dimensional embeddings.
We note that, although the method is not intrinsically restricted by input dimensionality, its performance still depends on constructing a meaningful neighbourhood graph, which can become challenging in high-dimensional settings.
The three-loop results are summarised in Figure 5 hyperlink. For embeddings and persistence diagrams for all three structures, see Appendix Figures A6–A11 hyperlink.
Response to Reviewer PLgL
We thank the reviewer for recognising the motivation behind TopoForce. Geometric distortion can make a loop hard to detect even when the underlying topology is unchanged. TopoForce uses multiscale diffusion on a neighbourhood graph to guide a low-dimensional layout that reveals structures at different scales.
Additional concern: Computational scalability
We agree that computational efficiency is important for practical applications. To improve the scalability of the layout optimisation, we implemented a Barnes--Hut approximation for repulsive force computation, which reduces computational cost by grouping distant points.
We evaluated this acceleration on four synthetic structures (circle, two-loop ring, three unequal loops and torus), keeping the graph, pairwise scales, initialization and iteration budget (1,600 steps) fixed across two seeds. At a Barnes--Hut threshold of 0.5, layout computation achieved a 4.1--6.2× speed-up, with aligned coordinate differences of only 0.84--3.53%. Importantly, identity-aware H1 matching on shared 350-point subsets retained all reference targets in the circle, ring and three-loop cases across both seeds. These results demonstrate that the layout stage can be substantially accelerated while maintaining the topological structures recovered by the original method. See Rebuttal Figure R3 hyperlink for the five-seed results.