Potential Flow Generator with $L_2$ Optimal Transport Regularity for Generative Models

George Em Karniadakis; Liu Yang

REVIEW 3 major objections 4 minor 1 cited by

Potential Flow Generator with $L_2$ Optimal Transport Regularity for Generative Models

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A potential flow generator can match a target distribution and approximate the L2 optimal transport map in a single pass.

desk verdict A clean and plausible way to graft L2 OT regularity onto arbitrary GAN losses, but the missing λ = 0 ablation leaves the key claim unproven. read the letter →

arxiv 1908.11462 v1 pith:LKSLARRA submitted 2019-08-29 cs.LG cs.CVstat.ML

Liu Yang , George Em Karniadakis This is my paper · ORCID

classification cs.LGcs.CVstat.ML

keywords potentialflowgeneratorL2optimaltransportHamilton-Jacobiequationgenerativeadversarialnetworkscontinuousnormalizingflowsunpairedimagetranslationphysics-informedneural

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper proposes a potential flow generator, a way to build the generator of a GAN or flow-based model from the flow of a gradient field $v=\nabla\varphi$ rather than from an arbitrary neural network map. The central claim is that this generator can simultaneously push the input distribution onto the target distribution and approximate the $L_2$ optimal transport map between them. The continuous version adds a penalty for the Hamilton–Jacobi residual $\partial_t\varphi+\frac{1}{2}|\nabla\varphi|^2=0$ to the original generator loss, and the paper argues that this penalty does not conflict with distribution matching. If correct, it gives a model-agnostic way to add optimal-transport regularity to many generative models and makes unpaired image translation behave like mapping each input to the nearest plausible output.

What carries the argument

The machinery is the Benamou–Brenier potential-flow formulation of optimal transport: among all density–velocity pairs $(\rho,v)$ satisfying the continuity equation, the $L_2$ transport cost is minimized when the velocity is a gradient $v=\nabla\varphi$ and the potential obeys the Hamilton–Jacobi equation $\partial_t\varphi+\frac{1}{2}|\nabla\varphi|^2=0$. The transport map is then obtained by integrating the ODE $du/dt=v(t,u)$. The paper's generators realize this by representing $\varphi$ with a neural network (continuous version, with the Hamilton–Jacobi residual penalized in the loss) or by time-discretizing the Hamilton–Jacobi step (discrete version). This formulation sidesteps the hard convexity constraint on the Brenier potential, since the Hamilton–Jacobi equation encodes optimality without an explicit convexity requirement.

What would settle it

Train the continuous potential flow generator on a pair of distributions whose $L_2$ optimal transport map is known analytically, scan the penalty weight $\lambda$ over $[0.1,10]$, and measure both the GAN loss and the error of the estimated map against the analytical map. If for any $\lambda$ in that range the distribution match degrades noticeably or the map error jumps well above the 0.02 level reported for the Gaussian and ring problems, the no-conflict claim is false.

Watch

Extended reading notes

Core claim

The paper's core discovery is that the optimal-transport structure of the $L_2$ Monge problem can be injected into a generative model by representing the generator as the time-$T$ flow of $v=\nabla\varphi$, where $\varphi$ solves the Hamilton–Jacobi equation $\partial_t\varphi+\frac{1}{2}|\nabla\varphi|^2=0$. In the continuous potential flow generator, $\varphi$ is a neural network and the Hamilton–Jacobi residual is penalized at points along sample trajectories; in the discrete version, the Hamilton–Jacobi step is encoded in the Euler update. The paper reports that the continuous version matches the target distribution while recovering the analytical optimal transport map in two 2D problems (Gaussian-to-Gaussian and ring-to-ring), with map errors around 0.02 and little sensitivity to the penalty weight $\lambda$ from 0.1 to 10. In unpaired image translation, the same construction maps digits 0–4 to similar-looking digits 5–9 and no-smiling faces to smiling faces while preserving identity, illustrating the 'proximity' property of $L_2$ optimal transport.

Load-bearing premise

The load-bearing premise is the paper's assertion in Section 3.2.2 that penalizing the Hamilton–Jacobi residual never conflicts with matching the generated distribution to the target; if the combined loss has spurious minima where the GAN loss is small but the map is far from optimal, the claimed robustness and correctness would fail.

Editorial extensions

If this is right

Any GAN whose generator loss is differentiable can be augmented with the continuous potential flow generator: replace the generator with the flow of $\nabla\varphi$ and add $\lambda$ times the squared Hamilton–Jacobi residual.
Flow-based models can use the same generator with maximum-likelihood training, because the likelihood ODE gives $\log p$ of outputs directly; the paper demonstrates this for three distribution pairs.
Unpaired image translation no longer needs a cycle-consistency pair of networks; a single potential flow generator maps inputs to nearby, semantically similar outputs.
Because results remain stable across $\lambda\in[0.1,10]$ in the reported experiments, users do not need to tune the PDE penalty weight carefully.
The discrete version encodes optimality without a penalty term but suffers from nested-gradient cost and occasional NaN, so the continuous version is the recommended default.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If the no-conflict claim holds generally, the Hamilton–Jacobi penalty acts as a selection rule: among the infinitely many maps that push $\mu$ to $\nu$, it picks one close to the $L_2$ optimal map, which would explain why it behaves like a regularizer rather than a competing objective.
The same potential-flow construction could be extended to other optimal-transport costs by replacing the Hamilton–Jacobi equation with the corresponding optimality PDE, although the paper does not test this.
A testable consequence of the 'proximity' property is that the learned map should have smaller expected squared displacement $\mathbb{E}\|G(x)-x\|^2$ than a GAN generator trained without the penalty; measuring this on held-out data would quantify the optimal-transport regularity directly.
The PCA embedding used for images may limit the method to modes captured by PCA; integrating the generator with autoencoder embeddings, which the paper mentions as future work, could sharpen the proximity property for high-frequency image details.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Here's the short version: the core idea is worth knowing, and the 2D evidence is decent, but the paper's most important claim—that the Hamilton–Jacobi penalty doesn't conflict with distribution matching—isn't isolated because there is no λ = 0 control for the continuous potential flow generator.

What's new: the authors represent the generator as the flow field v = ∇φ(t,x), with φ a neural network, and either discretize the HJ equation (discrete PFG) or penalize its residual (continuous PFG). This plugs into arbitrary GAN or flow-based losses, which distinguishes it from earlier OT-map methods like Seguy et al. or Leygonie et al. that require a specific loss. The 2D experiments on Gaussian and ring distributions compare against analytic OT maps and show errors around 0.02–0.03, with stability across λ = 0.1–10. That's real evidence the method does something.

The problems are empirical and structural. The comparison in Tables 1–2 is continuous PFG versus a vanilla MLP with an L2 transport penalty. That conflates architecture and penalty. The continuous PFG is a neural ODE with a constrained velocity field; that alone might produce near-OT maps in these simple problems. Without λ = 0 (or a tiny λ) for the continuous PFG, the robustness-to-λ claim doesn't show that the HJ penalty is doing the work. The 'no conflict' assertion in Section 3.2.2 is plausible for the exact problem—the OT map satisfies both objectives—but there's no analysis of the optimization landscape, and for finite-capacity networks a trade-off can reappear. The image translation results are qualitative, have no baselines, and use a crude PCA embedding; they don't carry much weight. No code or data is provided. The citations are appropriate, and the writing is clear.

If I were refereeing, I'd ask for the λ = 0 ablation first, then quantitative image metrics and code. The paper deserves a serious referee: the idea is testable, and the 2D evidence suggests it may hold up. The central claim is conditional, not refuted.

Referee Report

3 major / 4 minor

Summary. The paper proposes a 'potential flow generator' for generative models, in which the generator is the flow map of a time-dependent velocity field v = ∇φ(t,x). Two variants are introduced: a discrete version that encodes the Hamilton-Jacobi optimality condition through Euler time stepping, and a continuous version that represents φ(t,x) by a neural network and penalizes the Hamilton-Jacobi residual ∂tφ + 1/2|∇φ|^2 = 0 in the training loss. The generator is designed to be plugged into existing GAN and flow-based losses. The paper reports 2D experiments with known analytical optimal transport maps, showing smaller map errors than a vanilla generator with an explicit L2 transport penalty, and also reports image-translation demonstrations on PCA-embedded MNIST and CelebA data.

Significance. If the central claim is correct, the potential flow generator is a useful plug-in device: it aims to add L2 optimal transport regularity to arbitrary generative losses without the distribution-matching vs. transport-cost tradeoff that afflicts explicit L2 penalties. The paper's strengths include validation against analytical optimal transport maps in 2D (map errors around 0.018-0.031), a clean derivation of the Hamilton-Jacobi penalty from standard Benamou-Brenier theory, and the observation that the continuous formulation avoids the nested-gradient cost of the discrete formulation. However, the load-bearing 'no conflict' assertion is unproven, and the empirical evidence does not isolate the effect of the PDE penalty because no λ=0 ablation is reported for the continuous potential flow generator. The contribution is therefore promising but currently not fully substantiated.

major comments (3)

[Section 3.2.2, Eq. (16)] The assertion that 'there is no conflict between matching the distributions and penalizing Eqn. 7' is load-bearing but is not proved. For finite-capacity networks, exact simultaneous satisfaction of G#μ = ν and the Hamilton-Jacobi equation is not guaranteed, and the combined loss L_GAN + λ·PDE_residual could in principle have spurious minima in which the GAN loss is small but the map is far from optimal. The paper provides experiments but no analysis of the optimization landscape or approximation error. I request either a rigorous statement with explicit assumptions, or a concrete demonstration (e.g., tracking the PDE residual and distributional distance during training) that the tradeoff does not reappear.
[Section 4.1.1, Tables 1 and 2] The central robustness claim is not isolated because no λ=0 (or very small λ) baseline is reported for the continuous potential flow generator. The continuous PFG is a neural ODE with velocity constrained to v = ∇φ(t,x), a much more restricted map class than the vanilla MLP generator. It is possible that the near-optimal maps in Tables 1-2 are produced by this architectural constraint alone, rather than by the Hamilton-Jacobi penalty. Please report results for the continuous PFG with λ=0 (and, ideally, an intermediate λ value) in both 2D problems; without this ablation, the claim that the PDE penalty is the load-bearing component is not empirically established.
[Section 3.3.1, Eqs. (16)-(17)] The PDE penalty is evaluated only at sampled trajectory points of input samples, and the gradient of the loss with respect to the generator parameters is taken while treating the residual point coordinates as fixed. This means the Hamilton-Jacobi equation is enforced only on a low-dimensional set of trajectories, not on the full (t,x) domain. In high-dimensional tasks such as the PCA-embedded image experiments, there is no evidence that the penalty controls φ on off-trajectory points or on unseen inputs. Please justify this choice or add experiments with residual points sampled independently of the trajectories, and discuss the effect of this choice on the robustness claim.

minor comments (4)

[Section 4.1.1, Tables 1 and 2] The 'Error of map' metric is not defined in the text; please specify how the distance between the estimated and analytical transport maps is computed.
[Section 4.1.1, Table 2] The discrete PFG row is absent from Table 2 because of the reported 'NAN' problems; please state this explicitly in the table caption or note, since the reader otherwise cannot compare methods in Problem 2.
[Figure 2 caption] The caption says '(f-g) Continuous potential flow generator (PFG)' but the figure contains eight panels (a-h); please clarify which panels correspond to which λ values.
[General] There are several typographical errors: 'Eular' should be 'Euler' (Section 3.2.1), 'by by Chen et al.' should be 'by Chen et al.' (Section 3.3.2), and 'NAN' should be 'NaN' (Section 4.1.1).

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found; the L2 optimal transport regularity is imported from the external Benamou-Brenier theorem and checked against analytical maps.

full rationale

The paper's central derivation is not circular. The potential flow generator is built on the Benamou-Brenier characterization (Section 3.1), an external mathematical result, which states that the optimal velocity field for the L2 transport cost is v = ∇φ with ∂tφ + 1/2|∇φ|^2 = 0 (Eq. 7). The continuous PFG simply represents φ by a neural network and penalizes this HJ residual (Eq. 16 or 18), so the 'predicted' optimal transport maps are not fitted parameters renamed as predictions; they are the trained outputs of a network minimizing a loss whose ideal minimizer is the OT map by the cited theorem. The comparisons against the Gaussian analytic map f(x,y)=(2x,0.5y) and the ring map r -> r+1.5 in Tables 1-2 use externally derived references, not quantities used in training. The only self-citation is to the authors' PINN papers (Raissi et al. 2017a,b) for the generic idea of penalizing PDE residuals; this is not load-bearing because the HJ penalty is derived from Benamou-Brenier, not from the PINN papers. Two weaknesses should be noted as correctness risks, not circularity: (i) Section 3.2.2 asserts 'there is no conflict between matching the distributions and penalizing Eqn. 7' without proof, and the paper does not provide a λ=0 ablation isolating the effect of the HJ penalty; (ii) the robustness claim rests on a finite range of λ values in the experiments. These are empirical/analytical gaps, not instances where a claimed prediction reduces to its input by construction. Hence the circularity score is 0.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The central claim rests on standard optimal transport theory (Brenier, Benamou-Brenier), on the unproven assertion that the combined GAN-plus-PDE loss has no conflict between distribution matching and optimality, and on PINN-style approximation assumptions. No new physical entities are introduced.

free parameters (3)

PDE penalty weight λ = 0.1, 1.0, 10.0 tested in 2D; 1.0 in image tasks
Chosen by hand; balances distribution matching against Hamilton-Jacobi residual. The claimed robustness is demonstrated across this range, but the image tasks only use one value.
Time horizon T and number of steps n = T=1; n=4 (discrete), n=100 (continuous 2D), n=10 (image tasks)
Numerical discretization parameters for the flow ODE. The discrete version uses a very coarse n=4, which may limit its accuracy.
PCA embedding dimension = 100 (MNIST), 700 (CelebA)
Chosen by hand for image embedding; the blurriness of outputs is attributed to the low rank of PCA.

assumptions (6)

standard math Existence and uniqueness of the L2 optimal transport map as the gradient of a convex potential (Brenier's theorem).
Section 3.1 invokes Gangbo and McCann (1996) and McCann and Guillen (2011) for the Brenier potential form.
standard math The Benamou-Brenier dynamic formulation: minimal L2 transport cost equals the infimum of the action integral subject to the continuity equation, with optimality conditions v = ∇φ and ∂tφ + 1/2|∇φ|^2 = 0.
Section 3.1 cites Benamou and Brenier (2000) and uses Eqs. (4)-(7).
domain assumption The flow map generated by solving the ODE du/dt = v(t,u) from t=0 to T is a transport map from µ to ν when the initial distribution is µ and the terminal distribution matches ν.
Section 3.2 constructs the generator from the ODE (Eqs. 5, 10, 12); this assumes the velocity field is sufficiently smooth for a well-defined flow.
domain assumption A feed-forward neural network with tanh activation can approximate the potential φ and its gradients well enough on the support of the data.
Used throughout the experiments (Section 4); standard approximation assumption but not verified for the required accuracy.
ad hoc to paper Penalizing the HJ residual at sampled trajectory points enforces the PDE on the relevant domain.
The residual points are chosen on generated trajectories (Eq. 17); there is no error or convergence analysis for this PINN-style enforcement.
domain assumption The GAN or maximum likelihood loss drives G#µ to ν.
Standard in generative modeling; the paper does not analyze the interactions between the GAN loss and the PDE penalty.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Potential Flow Generator with $L_2$ Optimal Transport Regularity for Generative Models." pith.science (2026). https://pith.science/paper/LKSLARRA

@misc{pith2026190811462,
  author       = {Pith},
  title        = {Pith review of: Potential Flow Generator with $L_2$ Optimal Transport Regularity for Generative Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LKSLARRA}},
  note         = {Machine review of arXiv:1908.11462}
}

abstract

We propose a potential flow generator with $L_2$ optimal transport regularity, which can be easily integrated into a wide range of generative models including different versions of GANs and flow-based models. We show the correctness and robustness of the potential flow generator in several 2D problems, and illustrate the concept of "proximity" due to the $L_2$ optimal transport regularity. Subsequently, we demonstrate the effectiveness of the potential flow generator in image translation tasks with unpaired training data from the MNIST dataset and the CelebA dataset.

Figures

Figures reproduced from arXiv: 1908.11462 by the authors.

**Figure 1.** Schematic of generator without and with L2 optimal transport regularity. While both generators provide a scheme to map from the input distribution (purple) to the output distribution (orange), the total squared transport distances in the left generator is much larger than that in the right generator. Yang and Uhler [2018] proposed to learn the unbalanced optimal transport plan in an adversarial way derived from a co… view at source ↗

**Figure 2.** Comparison of different methods for problem 1. (a) Samples of [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗

**Figure 3.** Comparison of different methods for problem 2. (a) Samples of [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗

Figures from the paper (5 more)

**Figure 4.** Figure 4: Continuous potential flow generator in WGAN-GP for three different problems. Each row [PITH_FULL_IMAGE:figures/full_fig_p010_4.png]

**Figure 5.** Figure 5: Potential flow generator on the MNIST dataset. In each row, the top images are reconstructed [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]

**Figure 6.** Figure 6: Potential flow generator on the CelebA dataset. In each row, the top images are reconstructed [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]

**Figure 7.** Figure 7: We use a feed forward neural network with 5 hidden layers, each of width 128, to represent [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]

**Figure 7.** Figure 7: Potential flow generator in continuous normalizing flow for three different problems. Each [PITH_FULL_IMAGE:figures/full_fig_p013_7.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Optimal transport mapping via input convex neural networks
cs.LG 2019-08 conditional novelty 6.0 of 10

A principled minimax training procedure over input convex neural networks learns the optimal quadratic-cost transport map as the gradient of a convex potential.

Reference graph

Works this paper leans on

30 extracted references · 17 canonical work pages · cited by 1 Pith paper

[1]

Mart\' n Abadi, Ashish Agarwal, Paul Barham, Eugene Brevdo, Zhifeng Chen, Craig Citro, Greg S. Corrado, Andy Davis, Jeffrey Dean, Matthieu Devin, Sanjay Ghemawat, Ian Goodfellow, Andrew Harp, Geoffrey Irving, Michael Isard, Yangqing Jia, Rafal Jozefowicz, Lukasz Kaiser, Manjunath Kudlur, Josh Levenberg, Dandelion Man\' e , Rajat Monga, Sherry Moore, Derek...

2015
[2]

Gradient flows: in metric spaces and in the space of probability measures

Luigi Ambrosio, Nicola Gigli, and Giuseppe Savar \'e . Gradient flows: in metric spaces and in the space of probability measures. Springer Science & Business Media, 2008

2008
[3]

Wasserstein generative adversarial networks

Martin Arjovsky, Soumith Chintala, and L \'e on Bottou. Wasserstein generative adversarial networks. In International Conference on Machine Learning, pages 214--223, 2017

work page 2017
[4]

A computational fluid mechanics solution to the monge-kantorovich mass transfer problem

Jean-David Benamou and Yann Brenier. A computational fluid mechanics solution to the monge-kantorovich mass transfer problem. Numerische Mathematik, 84 0 (3): 0 375--393, 2000

2000
[5]

Neural ordinary differential equations

Tian Qi Chen, Yulia Rubanova, Jesse Bettencourt, and David K Duvenaud. Neural ordinary differential equations. In Advances in Neural Information Processing Systems, pages 6571--6583, 2018

work page 2018
[6]

Generative modeling using the sliced Wasserstein distance

Ishan Deshpande, Ziyu Zhang, and Alexander G Schwing. Generative modeling using the sliced Wasserstein distance. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 3483--3491, 2018

work page 2018
[7]

The geometry of optimal transportation

Wilfrid Gangbo and Robert J McCann. The geometry of optimal transportation. Acta Mathematica, 177 0 (2): 0 113--161, 1996

work page 1996
[8]

On a formula for the L _2 Wasserstein metric between measures on Euclidean and Hilbert spaces

Matthias Gelbrich. On a formula for the L _2 Wasserstein metric between measures on Euclidean and Hilbert spaces. Mathematische Nachrichten, 147 0 (1): 0 185--203, 1990

work page 1990

Show all 30 references

[9]

Generative adversarial nets

Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative adversarial nets. In Advances in Neural Information Processing Systems, pages 2672--2680, 2014

2014
[10]

Improved training of Wasserstein GANs

Ishaan Gulrajani, Faruk Ahmed, Martin Arjovsky, Vincent Dumoulin, and Aaron C Courville. Improved training of Wasserstein GANs . In Advances in Neural Information Processing Systems, pages 5767--5777, 2017

2017
[11]

Image-to-image translation with conditional adversarial networks

Phillip Isola, Jun-Yan Zhu, Tinghui Zhou, and Alexei A Efros. Image-to-image translation with conditional adversarial networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 1125--1134, 2017

2017
[12]

Principal component analysis

Ian Jolliffe. Principal component analysis. Springer, 2011

2011
[13]

Learning to discover cross-domain relations with generative adversarial networks

Taeksoo Kim, Moonsu Cha, Hyunsoo Kim, Jung Kwon Lee, and Jiwon Kim. Learning to discover cross-domain relations with generative adversarial networks. In Proceedings of the 34th International Conference on Machine Learning-Volume 70, pages 1857--1865. JMLR. org, 2017

2017
[14]

Adam: A method for stochastic optimization

Diederik P Kingma and Jimmy Ba. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980, 2014

2014 arXiv
[15]

Glow: Generative flow with invertible 1x1 convolutions

Durk P Kingma and Prafulla Dhariwal. Glow: Generative flow with invertible 1x1 convolutions. In Advances in Neural Information Processing Systems, pages 10215--10224, 2018

2018
[16]

MNIST handwritten digit database

Yann LeCun, Corinna Cortes, and CJ Burges. MNIST handwritten digit database. ATT Labs [Online]. Available: http://yann. lecun. com/exdb/mnist, 2, 2010

2010
[17]

Adversarial computation of optimal transport maps

Jacob Leygonie, Jennifer She, Amjad Almahairi, Sai Rajeswar, and Aaron Courville. Adversarial computation of optimal transport maps. arXiv preprint arXiv:1906.09691, 2019

1906 arXiv
[18]

Deep learning face attributes in the wild

Ziwei Liu, Ping Luo, Xiaogang Wang, and Xiaoou Tang. Deep learning face attributes in the wild. In Proceedings of the IEEE International Conference on Computer Vision, pages 3730--3738, 2015

2015
[19]

Five lectures on optimal transportation: geometry, regularity and applications

Robert J McCann and Nestor Guillen. Five lectures on optimal transportation: geometry, regularity and applications. Analysis and geometry of metric measure spaces: lecture notes of the s \'e minaire de Math \'e matiques Sup \'e rieure (SMS) Montr \'e al , pages 145--180, 2011

2011
[20]

Spectral normalization for generative adversarial networks

Takeru Miyato, Toshiki Kataoka, Masanori Koyama, and Yuichi Yoshida. Spectral normalization for generative adversarial networks. arXiv preprint arXiv:1802.05957, 2018

2018 arXiv
[21]

Viscous fingering: an optimal bound on the growth rate of the mixing zone

Felix Otto. Viscous fingering: an optimal bound on the growth rate of the mixing zone. SIAM Journal on Applied Mathematics, 57 0 (4): 0 982--990, 1997

1997
[22]

Physics informed deep learning (part i): Data-driven solutions of nonlinear partial differential equations

Maziar Raissi, Paris Perdikaris, and George Em Karniadakis. Physics informed deep learning (part i): Data-driven solutions of nonlinear partial differential equations. arXiv preprint arXiv:1711.10561, 2017 a

2017 arXiv
[23]

Physics informed deep learning (part ii): Data-driven discovery of nonlinear partial differential equations

Maziar Raissi, Paris Perdikaris, and George Em Karniadakis. Physics informed deep learning (part ii): Data-driven discovery of nonlinear partial differential equations. arXiv preprint arXiv:1711.10566, 2017 b

2017 arXiv
[24]

Variational inference with normalizing flows

Danilo Jimenez Rezende and Shakir Mohamed. Variational inference with normalizing flows. arXiv preprint arXiv:1505.05770, 2015

2015 arXiv
[25]

Improving GANs using optimal transport

Tim Salimans, Han Zhang, Alec Radford, and Dimitris Metaxas. Improving GANs using optimal transport. arXiv preprint arXiv:1803.05573, 2018

2018 arXiv
[26]

\ Euclidean, metric, and Wasserstein \ gradient flows: an overview

Filippo Santambrogio. \ Euclidean, metric, and Wasserstein \ gradient flows: an overview. Bulletin of Mathematical Sciences, 7 0 (1): 0 87--154, 2017

2017
[27]

Large-scale optimal transport and mapping estimation

Vivien Seguy, Bharath Bhushan Damodaran, R \'e mi Flamary, Nicolas Courty, Antoine Rolet, and Mathieu Blondel. Large-scale optimal transport and mapping estimation. arXiv preprint arXiv:1711.02283, 2017

2017 arXiv
[28]

Scalable unbalanced optimal transport using generative adversarial networks

Karren D Yang and Caroline Uhler. Scalable unbalanced optimal transport using generative adversarial networks. arXiv preprint arXiv:1810.11447, 2018

2018 arXiv
[29]

Dual GAN : Unsupervised dual learning for image-to-image translation

Zili Yi, Hao Zhang, Ping Tan, and Minglun Gong. Dual GAN : Unsupervised dual learning for image-to-image translation. In Proceedings of the IEEE International Conference on Computer Vision, pages 2849--2857, 2017

2017
[30]

Unpaired image-to-image translation using cycle-consistent adversarial networks

Jun-Yan Zhu, Taesung Park, Phillip Isola, and Alexei A Efros. Unpaired image-to-image translation using cycle-consistent adversarial networks. In Proceedings of the IEEE International Conference on Computer Vision, pages 2223--2232, 2017

2017

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

Mart\' n Abadi, Ashish Agarwal, Paul Barham, Eugene Brevdo, Zhifeng Chen, Craig Citro, Greg S. Corrado, Andy Davis, Jeffrey Dean, Matthieu Devin, Sanjay Ghemawat, Ian Goodfellow, Andrew Harp, Geoffrey Irving, Michael Isard, Yangqing Jia, Rafal Jozefowicz, Lukasz Kaiser, Manjunath Kudlur, Josh Levenberg, Dandelion Man\' e , Rajat Monga, Sherry Moore, Derek...

2015

[2] [2]

Gradient flows: in metric spaces and in the space of probability measures

Luigi Ambrosio, Nicola Gigli, and Giuseppe Savar \'e . Gradient flows: in metric spaces and in the space of probability measures. Springer Science & Business Media, 2008

2008

[3] [3]

Wasserstein generative adversarial networks

Martin Arjovsky, Soumith Chintala, and L \'e on Bottou. Wasserstein generative adversarial networks. In International Conference on Machine Learning, pages 214--223, 2017

work page 2017

[4] [4]

A computational fluid mechanics solution to the monge-kantorovich mass transfer problem

Jean-David Benamou and Yann Brenier. A computational fluid mechanics solution to the monge-kantorovich mass transfer problem. Numerische Mathematik, 84 0 (3): 0 375--393, 2000

2000

[5] [5]

Neural ordinary differential equations

Tian Qi Chen, Yulia Rubanova, Jesse Bettencourt, and David K Duvenaud. Neural ordinary differential equations. In Advances in Neural Information Processing Systems, pages 6571--6583, 2018

work page 2018

[6] [6]

Generative modeling using the sliced Wasserstein distance

Ishan Deshpande, Ziyu Zhang, and Alexander G Schwing. Generative modeling using the sliced Wasserstein distance. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 3483--3491, 2018

work page 2018

[7] [7]

The geometry of optimal transportation

Wilfrid Gangbo and Robert J McCann. The geometry of optimal transportation. Acta Mathematica, 177 0 (2): 0 113--161, 1996

work page 1996

[8] [8]

On a formula for the L _2 Wasserstein metric between measures on Euclidean and Hilbert spaces

Matthias Gelbrich. On a formula for the L _2 Wasserstein metric between measures on Euclidean and Hilbert spaces. Mathematische Nachrichten, 147 0 (1): 0 185--203, 1990

work page 1990