- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper matches the distribution itself of the Random Variable
$V= (X_1 - X_0 \mid X_t = x) $
to the field of mean vectors and variances (diagonalized). This differs from the usual Flow Matching that matches only the field to explicit samples from . The authors then utilize the learnt field for sampling of the usual ODE and also an instance of related SDE. Performance gain is probably from weighted ($gamma_{\theta, i}) flow matching objectives.
Critical strengths
Fitting a flow to vector field as a Random Variable is a worthy hindsight.
Critical weaknesses
Weaknesses
This paper is badly written with no regard to the reader's structural mathematical understanding. Most notably, it does not clearly distinguish what's random and what is not. Lowercase is once introcued as constant in equation (2), so considering its law is nonsensical (12) if we read this literally--- requiring the reader to reconstruct what is what from the context. But after a long while, I realized that the core of this paper is distributional matching of
to a model.
Such restructuring of consistency shall not be the reader's task. The document as a whole required the readers to restructure the logic.
Also, I understand the analogy of Section 3.3 as a separate story, but if one comes to apply a from a different model to this, it is a big leap in concept that requires more justification. The system of SDE is governed with a completely different filtration, and while the result is interesting, it appears to be an "additional technical attempt" to be in the discussion—it added more to my confusion.
I support the idea (after I understood it) and support the result, but I don't recommend the paper of this construction for publication, and I feel that it needs major revision as well as additional theoretical work for connecting what seems to be all over the place.
Questions for authors
Flow matching and Diffusion models can both be understood in one common framework as Gradient Flow (See, https://arxiv.org/pdf/2204.04853(ICLR2023) or (https://arxiv.org/pdf/2407.04495 Physical Review X ), for few of examples, and they all differ only through what energy functional they are associated with:
What is the Energy functional does this poseterior Flow correspond to?
Why is this paper's useful, and when is it expected to be good choice and when it is not? I feel that the explanation of it shall be better because it is probabilsitic is insufficient, since the preferred choice of probabilistic model as well as preferred choice of entropy differs across situations--- with myriads of tricks being found today, I believe that each researcher is responsible for answering such questions.
Likewise, when this is used for the related diffusion model (19), how does it change the original diffusion's energy functional?
In a related note, I feel that the performance gain is probably arising from weighted ($gamma_{\theta, i}) flow matching objectives. Please further elaborate in relation to flow-theory.
6 comments
2/4 — The reference energy and Eq. 19
PFM and FM use the same reference velocity, so they share the functional in your equation. Let be the density of , where is standard Gaussian and independent of the data endpoint . On , take positive smooth densities with finite displayed moments and entropies and the stated integration conditions. The velocity-score identity in submission Eq. 8, derived in Appendix A.4, gives
In the sign convention , take :
The first variation is . Evaluated at , its spatial gradient is , so the equation gives the continuity equation for the reference path. The usual Wasserstein descent convention uses . This functional represents the transport along the prescribed reference path. Eq. 13 gives the loss used to train PFM.
Koshizuka and Sato [1, Secs. 3.1–3.2] use a Lagrangian action between population marginals. Ikeda et al. [2, Secs. III–IV] relate diffusion accuracy to entropy production and Wasserstein geometry.
For score-based diffusion, the probability-flow ODE and the reverse SDE with tensor share the noising-path marginals with the exact score and matching initialization (submission Eqs. 5-6 and Appendix A.6). Eq. 19 applies the ODE-SDE correspondence in Eq. 7 to the FM path , with . With the reference path and its velocity fixed, choosing a different tensor and applying its drift correction leaves the reference functional above unchanged. It changes how perturbations dissipate. Appendix A.5 gives the conditions for matching the one-time marginals, and Proposition 4.2 and Appendix C.3 give the relative-entropy calculation.
References
-
[1] Koshizuka and Sato. Neural Lagrangian Schrödinger Bridge: Diffusion Modeling for Population Dynamics. ICLR, 2023.
-
[2] Ikeda et al. Speed-accuracy relations for diffusion models: Wisdom from nonequilibrium thermodynamics and optimal transport. Physical Review X, 15, 031031, 2025.
3/4 — Why use the posterior covariance?
The learned covariance estimates how much the velocity targets vary in each coordinate at a given state and time. We choose the SDE noise so that it is isotropic after dividing by the posterior standard deviations (Eqs. 17-18 and Appendix C.1):
The learned covariance sets the relative noise strengths across states and coordinates, and sets the overall strength. With the exact mean and derived score, the score and divergence corrections in Eq. 19 preserve the reference marginals under the conditions in Appendix A.5.
To see how this noise affects a density perturbation, let be the reference density and evolve under the same SDE from a possibly different initial law. The drift for both uses the score of . Appendix C.3 gives the equation for (Eq. 49):
The derivative of accounts for the motion of both densities. Their common transport terms cancel. For positive smooth densities, finite integrals, and vanishing boundary terms, Proposition 4.2 gives
Each squared coordinate derivative in the integral is weighted by the learned conditional velocity variance at that state, and scales the overall dissipation rate. With throughout, both densities follow the same ODE and their relative entropy stays constant. Appendix H.6 illustrates the calculation with exact coefficients.
In submission Figure 7, we vary the overall noise strength and corresponding drift correction with the learned covariance field fixed. Moderate stochasticity improves generation over PFM ODE, while the largest tested strength reverses these gains. This supports choosing the overall strength for the task and step count without retraining. A covariance that is proportional to the identity and depends only on time gives an isotropic sampler. The ImageNet-256 comparisons below test local covariance at matched sampling time, including the cost of variance derivatives (revised Sec. 5.3 and Appendix H.7).
Table 1. ImageNet-256 at a fixed PFM checkpoint
| Sampler | 250-step FID | Equal-time FID |
|---|---|---|
| PFM linear-schedule isotropic SDE | 23.63 | 23.19 |
| PFM covariance-fitted isotropic SDE | 23.31 | 23.02 |
| PFM SDE | 22.34 | 22.34 |
Each coefficient is selected independently with the same validation budget. At 4.01-4.02 ms/sample, the samplers use 283/282/250 steps in the order shown.
The linear-schedule isotropic SDE uses a prescribed scalar schedule. The covariance-fitted isotropic SDE multiplies the same linear time factor by the average covariance trace per dimension and tunes its own coefficient. Each SDE uses the drift correction for its tensor, and the two isotropic tensors have zero spatial divergence. PFM SDE has the lowest FID at 250 steps and at matched sampling time.
1/4 — The velocity target and the sampler's filtration
Thank you for pointing out where the notation made the argument hard to follow. The revised paper distinguishes the reference variables, the moments PFM learns, and the sampler that uses them.
The v in Eq. 2 is a conditional mean. The random velocity target is V=X1−X0. Here X0∼p0 and X1∼pdata are independent, and Xt=(1−t)X0+tX1 (Eq. 1). At a fixed (x,t), v(x,t)=E[V∣Xt=x] is a deterministic vector and vθ(x,t) estimates it. We rewrote Eq. 12 in Sec. 4.1 to show that the Gaussian approximates the conditional law of V:
qθ(⋅∣x,t)=N(vθ(x,t),Σθ(x,t)).
At the population optimum, the Gaussian fits the conditional mean and componentwise variances when they are attainable (Proposition 4.1 and Appendix B.4). For a fixed inexact mean, the pointwise variance target is the conditional residual second moment, clipped to the stated bounds (Eqs. 43-44). The affine relation in Eq. 10 gives the endpoint moments in Eq. 15:
μθ=x+(1−t)vθ,Σθx=(1−t)2Σθ.
The SDE evaluates the learned fields at its own state Yt. Once training is finished, the drift and diffusion coefficients are fixed functions of state and time. The sampler's filtration includes its initial state and Brownian history. Since Yt is adapted to this filtration, the coefficients evaluated at Yt are adapted as well (revised Appendix A.5).
The original Sec. 3.3 and Appendix A.5 explain why the sampler can have the reference marginals. Let pt be the density of Xt. Eq. 8 takes X0∼N(0,I) to relate the velocity to the score. With the exact velocity and score, Eq. 19 uses aθ=λΣθ and bθ=v+(∇⋅aθ+aθ∇logpt)/2. The product rule gives
−∇⋅(bθpt)+21∑i,j∂i∂j(aθ,ijpt)=−∇⋅(vpt).
With the same initial law, regularity, boundary conditions, and unique density evolution as in Appendix A.5, Yt has density pt at each time. The calculation works for any admissible diffusion tensor with its corresponding drift correction. PFM uses the learned covariance to set that tensor.
The reference process and the sampler can have different joint laws across time. At a visited state Yt=y, Eq. 15 estimates the mean and covariance of the reference endpoint posterior p(X1∣Xt=y) (original Sec. 4.3 and Appendix D.3). The sampler's own conditional endpoint law comes from its transition dynamics. The one-time marginal result concerns the exact continuous SDE. Appendix G.4 analyzes the additional randomness from estimating the derivative in the finite-step sampler.