- Rating
- 4
- Confidence
- 2
- Informativeness
- 3
- Presentation
- 2
- Soundness
- 3
Read review
Paper summary
The paper introduces the Cartan-Hadamard Sliced-Wasserstein (CHSW) manifold distance for fair representation learning that takes into account that geodesic distances differ from the usually chosen euclidean assumption. It is theoretically shown that minimizing this distance actually bounds fairness scores (i.e. demographic parity) and demonstrates empirical advantages on selected (theoretical and real-world) fairness benchmarks.
Critical strengths
-
The work is well motivated, well written and theoretical strong. The idea of extending fair representation learning to take the manifold structure into account is relevant and interesting. In particular, the paper addresses a well thought-of research question throughout.
-
The theory shows that the proposed distance can be used to bound demographic parity for pullback euclidean manifolds and also introduces a soft DP violation for more general Hadamard manifolds also including a bound by the proposed manifold distance.
Critical weaknesses
First
On such a manifold, Euclidean distance is the wrong ruler: it measures separation along straight lines through the ambient space, whereas a downstream predictor, or an adversary trying to recover the protected attribute, operates in the distance along the manifold
But is this really the case in Fair Representation Learning? Usually the manifold hypothesis directly affects the original input space of the data, but it is also known that Deep Learning usually disentangles this structure (e.g. [1]) so this provides evidence for the contrary, e.g. that assuming an euclidean representation of the embedding space and relying on an euclidean penalty should be fine. While Figure 6 and 7 only test the network size for the proposed approach it would be interesting to see if the same conclusion hold for any other baseline (i.e. [1] suggests that manifold distances matter less the bigger the network), i.e. currently large networks seem to not perform well for the proposed approach but it would be relevant to see if larger networks are more beneficial for the baselines.
Currently the main motivational figures are in the appendix, it would be further helpful to include these in the main text and also show how the proposed CHSW is computed in contrast to the euclidean approach.
Second
Whilst the results in Table 1 favor the proposed approach, it is not discussed that in several cases the "best performance" comes with only marginal improvements yet much higher standard deviation, which is also not discussed in section 5.2.1 (for example on SPD and Pullback for change in demographic parity).
Further, Figure 3 does not report standard deviation and is missing a "dummy" baseline, i.e. a constant predictor to anchor the results (i.e. the results in the bottom right probably just correspond to a constant model that predicts the trainings data mean). Both changes are highly helpful in actually interpreting the results.
Also the authors report to test coefficients up to which is super confusing as such extreme penalties should not be useful and numerically extremely unstable (and this should probably be evident from the standard deviation).
To maintain my already positive score it would be required to incorporate the changes suggested in point 2 where for point 1 it is sufficient to adapt the framing accordingly or add convincing arguments why manifold distances are still required (e.g. based on the empirical results later).
[1] Why Deep Learning Works: A Manifold Disentanglement Perspective, Brahma et. al. (IEEE TNNLS, 2016)
Questions for authors
-
Is there a reason you choose or expect such values to be useful?
-
Could you add a notation section ? This would make reading the paper much more easy, even though many seem to follow standard convention it takes extra effort to confirm, e.g. I assume that indicates the tangential plane touching origin and that is the set of all probability measures on ?
-
I assume that the approach works for any discrete groups and arbitrary targets (i.e. also for regression) but not for continuous sensitive attributes right?
6 comments
4. Distance diagnostics across network sizes. We additionally conducted a separate diagnostic using CHSW models trained with a fixed across the same 16 architecture settings. For each model, we sampled 2000 distinct test pairs and compared hyperbolic geodesic distances between mapped representations with two Euclidean distances: the ambient distances between the mapped manifold points, and the distances between the raw encoder outputs. The result is shown in figure 4(b) in page 10.
For the largest network, , the corresponding Spearman correlations are and , respectively. Thus, even for this large network, Euclidean distances in the raw encoder coordinates do not closely preserve the distance rankings induced by the specified hyperbolic geometry. Meanwhile, ambient Euclidean distances between mapped points closely preserve these rankings.
5. Main-text motivation and computational comparison. We have moved the motivating example (former Figures 1 and 4) into Section 1 of the main text (Figure 1, page 2), with additional Euclidean SW, MMD, and Sinkhorn values. The example shows that small ambient Euclidean discrepancies can coexist with substantial separation under the hyperbolic geometry, clarifying why Euclidean coordinate distances may not faithfully reflect intrinsic geometry when representations are modeled on curved manifolds.
The computational distinction is as follows. Euclidean SW uses linear projections of encoder outputs, whereas CHSW uses geometry-aware one-dimensional projections on the representation manifold. Both methods then compute Wasserstein barycenter discrepancies using empirical quantiles and average over slicing directions. We have added this computational comparison to Appendix E.1 (pages 26–28).
Q2: Discussion of improvement magnitudes and variability on Pullback and SPD.
A: Thank you for highlighting this. We agree that the original Section 5.2.1 insufficiently discussed effect sizes and variability. On Pullback and SPD, mean improves from to and from to , while the corresponding standard deviations increase from to and from to . These are modest mean improvements relative to the across-seed variability. We have revised Section 5.2.1 (pages 7-8) to explicitly discuss these numerical comparisons and interpret the mean improvements alongside their variability.
Fairness should also be assessed jointly with accuracy. For example, on SPD, CHSW achieves mean accuracy of , compared with for , the second-best method in mean hard DP, while also reducing mean DP. Figure 3 provides complementary trade-off evidence, particularly for soft DP and on SPD. The revised discussion therefore emphasizes the favorable fairness–accuracy balance while acknowledging the modest improvements and higher variability in these cases. All changes are highlighted in red.
Q3: Variability estimates and a constant-predictor reference for Figure 3.
A: Thank you for these helpful suggestions. The original Figure 3 showed single-seed curves without variability estimates. We have revised Figure 3 on page 9 to report mean curves with error bars indicating standard deviation across seeds, and added the suggested dummy baseline—a constant predictor that outputs the mean training label. The revisions are highlighted in red.
Q4: Is there a reason you choose or expect such values to be useful?
A: Thank you for pointing out this typo. The intended value is , not . We have corrected it in Section 5.4 on page 9 and highlighted the correction in red.
Q5: Notation and symbol definitions.
A: Thank you for this helpful suggestion. We have added a notation table (Table 5, “Main notation used in this paper”) in Appendix B (pages 15–17). In particular, denotes the tangent space at the reference point , and denotes the set of Borel probability measures on .
Q6: Applicability to discrete groups, regression, and continuous sensitive attributes.
A: Our regularizer supports finitely many discrete sensitive groups and is independent of the prediction target. It can therefore be combined with a regression loss, although our stated fairness guarantees concern classification scores and decisions.
Continuous sensitive attributes are not directly covered by the current formulation. Extending the approach would require adapting the conditional-distribution estimation and regularization, together with further theoretical analysis. We focus on the widely adopted discrete-group setting and regard continuous attributes as an interesting extension.
report mean curves with error bars indicating standard deviation across seeds
Thank you, visualizing them with a shaded region would be preferable for readability.
Our regularizer supports finitely many discrete sensitive groups and is independent of the prediction target.
Can you make that more clear with a short sentence in the beginning of the paper ? This is otherwise a common point of confusion in the literature thus it would be helpful for the community.
Thank you for pointing out this typo. The intended value is , not . We have corrected it in Section 5.4 on page 9 and highlighted the correction in red.
This is a bit unexpected: your provided code quite clearly states "lambdas=[0,1,10,100,1000,10000,100000,10000000000,100000000000000]", are your results now run with these lambdas or your original values and which values do the paper results correspond to?
We sincerely thank the reviewer for the thoughtful review and encouraging assessment of our work. We particularly appreciate your recognition of the relevance of manifold-aware fair representation learning, the clarity of our presentation, and the strength of our theoretical analysis. Your positive assessment of both the demographic-parity bound on pullback Euclidean manifolds and the soft demographic-parity bound on general Hadamard manifolds is especially encouraging. Below, we address your comments and suggestions to further improve the paper. All changes in the revised manuscript are highlighted in red.
Q1: Representation geometry, baseline capacity comparisons, main-text motivation, and computational illustrations.
A: Thank you for this thoughtful comment. We agree that the quoted sentence is stated too strongly, and we have revised it in Section 1 (pages 1-2). Our premise concerns the representation space on which the fairness regularizer acts, not the input space. Below, we clarify this motivation through related work and provide additional experiments examining network capacity and learned representation geometry.
1. Deep networks do not necessarily flatten their representations. Deep networks can disentangle features and improve class separability without necessarily making the representation geometry flat. Ansuini et al. [1] find that learned representations can remain curved even in the last hidden layer, suggesting that flattening is not a general outcome of deep representation learning. Kaufman and Azencot [2] directly study latent image manifolds and observe substantial curvature in the final layers of trained networks across architectures. Khrulkov et al. [3] report high hyperbolicity of image datasets measured through features extracted by pretrained networks, motivating the use of hyperbolic image representations. Together, these studies support examining representation geometry empirically rather than assuming that increasing network capacity necessarily makes Euclidean distances sufficient.
[1] Ansuini A et al. Intrinsic dimension of data representations in deep neural networks. NeurIPS, 2019.
[2] Kaufman I et al. Data representations’ study of latent image manifolds. ICML, 2023.
[3] Khrulkov V et al. Hyperbolic image embeddings. CVPR, 2020.
2. Group information can persist despite apparent invariance under a restricted criterion. Elazar and Goldberg [4] report that, in their experiments, an adversarial component can reach chance-level accuracy during training while a post-hoc classifier still recovers demographic information above chance. Ravfogel et al. [5] show that guardedness against log-linear prediction does not generally prevent a multiclass downstream log-linear model from indirectly recovering the erased concept. These findings motivate distinguishing successful task prediction or restricted invariance from full removal of group information. They complement the geometric motivation, rather than directly establishing that Euclidean penalties are inadequate.
Our framework accommodates both flat and curved representations. For manifold-valued encoders, such as hyperbolic embeddings or SPD matrices, CHSW measures distributional discrepancies using the specified representation geometry. For pullback Euclidean manifolds, CHSW reduces to ordinary sliced Wasserstein distance in the transformed coordinates (Appendix F.3.4). Consistent with this flexibility, manifold selection yields pullback Euclidean geometry for Adult and hyperbolic geometry for LFWA. We therefore treat geometry as a modeling choice evaluated through predictive and fairness performance, rather than assuming that Euclidean regularization always fails.
[4] Elazar Y et al. Adversarial removal of demographic attributes from text data. EMNLP, 2018.
[5] Ravfogel S et al. Log-linear guardedness and its implications. ACL, 2023.
3. Capacity comparisons with all five baselines. Following the suggestion, we extended the LFWA network-size analysis to all five baselines, using the same data splits and the full grid dz,h∈ {16,32,64,128 }. We explored regularization coefficients separately for different methods and architecture settings, since a single coefficient need not be appropriate for every network size. The supplementary figure4(a) in page 10 report ACC, ΔDP, and ΔDPsoft.
Larger networks can benefit the baselines, but CHSW also performs well at large capacities. For example, at dz=h=128, CHSW achieves ACC of 91.29%, hard DP of 5.68%, and soft DP of 5.15%, compared with 91.06%, 6.98%, and 7.07%, respectively, for SW. Thus, the observed behavior is not that CHSW necessarily deteriorates as the network becomes larger.