- Rating
- 1
- Confidence
- 3
- Informativeness
- 1
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper extends flow matching methods from Gaussian distributions to exponential power distributions. They derive the conditional flows using Fisher-Rao geodesics rather than the distance defined in the input space. The method is evaluated on synthetic datasets, MNIST, and QM9.
Critical strengths
The idea of using the Fisher-Rao geodesic is actually interesting, but I'm not fully convinced why evaluate the method on these Euclidean datasets.
Critical weaknesses
Critical Weaknesses
-
Theoretical link between conditional Fisher Rao optimality and marginal generation: Theorem 4.1 provides a very clean result in that the proposed schedule is in fact the Fisher Rao geodesic between the conditional endpoints. In practice however the PG path acts as an affine interpolant with a schedule that is sample dependent in that it is determined by . It's not at all clear from a theoretical point of view why Fisher Rao optimality at the conditional level should necessarily lead to a more learnable marginal vector field or better sampling.
-
The empirical performance of the pure PG path: The geometric formulation of PG is by far the most significant conceptual contribution, but the pure PG path has yet to show a consistent empirical advantage over the standard OT baseline across the benchmarks reported. Specifically:
- In Table 2 the isotropic PG variant is worse than OT on both ED and MMD across both settings.
- In Tables 3 and B.4 pure PG performs worse than OT on all three evaluation metrics for both values.
- In Table 4 pure PG gives higher FID and classifier FID scores than OT for both settings.
-
Source of empirical gains: The majority of the empirical gains are attributable to the anisotropic version or to the hybrid (HB) approach which switches over to the standard OT path at . A number of the improvements are actually within one std of the baseline, so it's not easy to say that the gains are due to the Fisher Rao geodesic construction per se.
-
Alignment between the text claims and the tabulated results: There are several instances where the summary statements in the main text don't match the numbers in the tables. Line 461 states that HB achieves the lowest and ED at , yet in Table 2 HB's is actually higher than a number of the baselines and the anisotropic PG variant has a lower ED. The generalisation that almost universally outperforms is also contradicted by multiple examples in Tables 2, 3, B.3 and B.4.
-
Baseline consistency on Euclidean data: On Euclidean data with a flat metric RFM reduces exactly to standard OT flow matching. On Table 2 (at ) there is a significant difference between the two methods ( of 5.49 vs 5.66, with reported std devs of around 0.01). The difference between these two baselines which should be identical is on a similar order to the improvements seen from the proposed variants so that does raise some questions about variance and baseline calibration.
Questions for authors
- What result can show that minimizing Fisher-Rao energy can improve generation, rather than simply defining a different interpolation in another space?
- How are the two branches of HB joined at ? Is there some matching step in the implementation, and how is the training target kept consistent when sampling with an ODE?
- How does the experimental part compare with RFM? What manifold / metrics and conditional path are used on these Euclidean datasets?
Correction of two columns misplaced in Table 5
In response to your comment on "QM9, PG’s validity (90.63%) falls below all four external baselines (94.4–98.9%), a gap described as “slightly lower” (Sec. 5.3). ", we realized we misplaced results of OT/PG/HB in two columns Mols. Val. () and Mols. S. () in the submission and have fixed them by swapping them back (marked in red in the revision) -- note that the percentage of valid molecules (Mols Val) should be greater than or equal to that of stable molecules (Mols. S), i.e. Mols. Val. () ≥ Mols. S. ().