Pith. sign in

REVIEW 4 major objections 5 minor 39 references

Non-Normal Diffusion Models

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read This paper claims that diffusion models keep the same continuous-time limit no matter the distribution of their noise steps, provided each step has the matching mean and variance.

desk verdict The invariance idea is right but not new, and the loss derivations rest on an invalid ELBO; the paper needs major revision before it contributes. read the letter →

arxiv 2412.07935 v1 pith:LOUOBERG submitted 2024-12-10 cs.LG cs.AI

Henry Li This is my paper · ORCID
classification cs.LGcs.AI MSC 60F1760H1068T07
keywords diffusionmodelsnon-Gaussianincrementsinvarianceprinciplestructuredrandomwalksscore-basedgenerativelikelihoodboundmodeling
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

Diffusion models normally inject Gaussian noise at every step, and the Gaussian assumption is treated as part of the model. This paper tries to establish that the Gaussian is not needed: as the step size goes to zero, any increment distribution with the correct conditional mean and variance drives the same continuous-time diffusion. If the claim holds, the noise distribution becomes a free design parameter, and diffusion models gain new families of training losses and new visual behavior. The paper proves an invariance principle for structured random walks, derives four concrete non-Gaussian diffusion models, and tests them on image data.

What carries the argument

The central object is the structured random walk, whose increments decompose as $$\$\Delta$ x_k = f(x_k,t_k)\,\$\Delta$ t_k + g(t_k)\sqrt{\$\Delta$ t_k}\,z_k$$ with $\mathbb{E}[z_k]=0$ and $\operatorname{Var}(z_k)=1$, plus a finite-fourth-moment condition on $z_k$. Theorem 3.2 is the load-bearing result: it supplies the invariance that lets any such zero-mean, unit-variance noise define a valid diffusion model in the small-step limit. Closed-form KL divergences among Gaussian, Laplace, and uniform increment families then translate that invariance into concrete loss functions.

What would settle it

Take a simple Ornstein-Uhlenbeck SDE and simulate structured random walks with $z_k$ drawn from Gaussian, Laplace, uniform, and a two-point distribution with matching mean and variance; if the finite-dimensional distributions at a fixed time $t$ do not all converge to the same OU Gaussian law as $\Delta t_k \to 0$, Theorem 3.2 is false. Separately, one can check Eq. (5) directly: for a Laplace-increment model with a fixed score network, compute the left- and right-hand sides by Monte Carlo on a small dataset and see whether the claimed lower bound actually holds.

Watch

Extended reading notes

Core claim

The paper's central claim is Theorem 3.2, the Structured Invariance Principle. It states that, under mild regularity conditions (Lipschitz drift and diffusion, linear growth, and finite fourth moments of the noise variables), the càdlàg continuous-time interpolation of any structured random walk converges in distribution to the solution of the corresponding Itô SDE as the time steps go to zero. Because a structured random walk only pins down the first two conditional moments of each increment, the theorem says any zero-mean, unit-variance noise source works in the limit. The paper uses this to argue that the reversed diffusion process, and therefore the likelihood bound and training losses, is invariant to the distribution of the increment $\Delta x_k$ when the step size is small, so Gaussian increments are a choice, not a requirement.

Load-bearing premise

All the proposed non-Gaussian loss functions inherit their validity from the likelihood bound in Eq. (5), which is asserted without derivation; if that bound is not a genuine evidence lower bound for arbitrary increment distributions, the new training objectives have no theoretical justification.

Editorial extensions

If this is right

  • In the small-step limit, the choice of increment family $\{z_k\}$ --- Gaussian, Laplace, uniform, or other --- does not change the continuous-time process being sampled.
  • New training objectives become available: a Laplace/Laplace model yields a loss that behaves like a weighted $\ell^1$ error on the score residual, while uniform/Laplace pairs produce a piecewise $\ell^2$/ $\ell^1$ loss.
  • Diffusion models can be designed by picking the increment distribution separately from the score-network parameterization, so the noise family is a tunable stylistic and inductive-bias knob.
  • The experimental results indicate that non-Gaussian increments retain competitive likelihood and sample quality while producing visually different samples, such as more saturated colors for Laplace increments.

Reading between the lines

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

  • Beyond the paper: the same invariance suggests that noise distributions with finite variance but non-trivial higher moments could inject prior structure into generation, for example sparse noise for sparse data, although the paper only demonstrates Laplace and uniform cases.
  • Beyond the paper: because the proof relies on finite fourth moments, heavy-tailed increment distributions with infinite variance would be expected to break convergence; testing a Cauchy-increment random walk would directly delineate the boundary of the theorem.
  • Beyond the paper: the framework implies that the standard score-matching loss is only one point in a continuum of objectives, so a practical next step is to benchmark whether the $\ell^1$-type losses improve robustness to outlier pixels or adversarial perturbations in generated images.
  • The paper leaves the statistical guarantees of these alternative objectives open; verifying consistency and asymptotic normality for the non-Gaussian losses would determine whether they are sound estimators beyond their empirical performance.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes to generalize diffusion models by allowing the step increments Δx_k to be drawn from non-Gaussian distributions. Its main theoretical contribution is Theorem 3.2, a 'structured invariance principle' stating that a random walk with increments Δx_k = f(x_k,t_k)Δt_k + g(t_k)√Δt_k z_k, where z_k has zero mean and unit variance, converges in distribution to the Itô diffusion (1) as the step size goes to zero. On this basis the paper derives several training losses (Eqs. 9, 10, 13, 14) corresponding to Laplace and Uniform choices for q and p_θ, and reports CIFAR-10 experiments in Table 2 claiming competitive likelihood and sample quality.

Significance. If Theorem 3.2 were correctly proved and the likelihood bound (5) were valid, the paper would identify a genuinely useful design degree of freedom: the finite-step increment distribution in diffusion models. The proposed losses are explicit and are stated in a way that is directly testable, and the KL computations in Appendix A.1–A.3 are checkable. However, the central derivation is not sound. The ELBO in Eq. (5) is asserted without proof and is not a valid variational bound for non-Gaussian increments, and the tightness argument for Theorem 3.2 contains a concrete mathematical error. Because the training losses in Section 4 all depend on Eq. (5), the central claim of a generalized training framework is unsupported as it stands. The contribution is therefore conditional on a substantive reworking of the theory rather than on local corrections.

major comments (4)
  1. [Section 2, Eq. (5); Appendix A.5] Equation (5) is the likelihood bound from which every loss in Section 4 is derived, but no derivation is given and, as written, it is not a valid variational lower bound for the reverse model in Eq. (4). A valid ELBO must involve a variational family over the latent trajectory conditioned on x0; in standard diffusion models this is q(x_{k-1} | x_k, x_0), whose dependence on x0 is essential. The paper instead writes KL(q(Δx_k | x_{k+1}) || p_θ(Δx_k | x_{k+1})) with no conditioning on x0 and no explicit variational family. For non-Gaussian structured increments, q(Δx_k | x_{k+1}) is not determined by the first two moments and is generally intractable. Moreover, the appendix derivations in Lemmas A.1–A.4 do not evaluate the expression in Eq. (5): they compute KL divergences between forward transition densities conditioned on x_k, e.g., p_θ(Δx_k | x_k) = N(Δx_k; f̂_θ(x_k,t_k)Δt_k, g(t_k)^2 Δt_k), rather than the reverse kernels q(Δx_k | x_{k+1}) and p_θ(Δx_k | x_{k+1}) appearing in Eq. (4). Therefore Eqs. (9), (10), (13), and (14) are not justified as ELBO-based losses. This gap is load-bearing: without Eq. (5), the paper's central claim of a generalized training framework is unsupported, and Theorem 3.2, even if repaired, concerns only the forward process and does not repair this issue.
  2. [Appendix B.3, Lemma B.1, Eq. (68)] The tightness argument in Lemma B.1 does not establish the Kolmogorov criterion it invokes. Equation (68) gives an upper bound E||Δx_k||^4 ≤ C√Δt (the displayed constant is O(√Δt)), but the preceding line actually suggests a bound of order Δt for the drift term and √Δt for the diffusion term, so the claimed order is already not justified. More importantly, in Case 1 the paper derives E||x_n(s) − x_n(t)||^4 ≤ (n|t−s|) E||Δx_k||^4 ≤ C_1 √n |t−s|, and then, using |t−s| ≤ n^{−1}, asserts a |t−s|^2 bound. This is not a passage from a first moment to a fourth power; it is already the fourth moment, and the bound is O(n^{−1/2}), not O(|t−s|^2). Case 2 has the same defect. Case 3 relies on Lemma B.3, which is stated for sums of independent mean-zero terms and is not connected to the original dependent increments in a way that yields the required |t−s|^2 bound. Thus the tightness half of Theorem 3.2 is not proved.
  3. [Theorem 3.2 and Section 2] Even if Theorem 3.2 were proved, it would not establish the abstract's claim that the reversed process is invariant to the distribution of the increments. The theorem is a forward-time weak convergence statement for the càdlàg interpolation of the structured random walk. The reverse-time generative model in Eqs. (2)–(4) requires convergence of the one-step reverse conditionals q(x_k | x_{k+1}) (or q(Δx_k | x_{k+1})) to the reverse transitions of the limiting SDE. Weak convergence of forward paths does not imply convergence of these reverse conditionals for non-Gaussian increments; the reverse conditionals depend on the full marginal density of x_{k+1}, not just on f, g, and the first two moments of z_k. No lemma in Appendix B addresses reverse-time conditioning. Consequently the paper's main conceptual claim—that the reversed process is invariant to the increment distribution—is formally unsupported.
  4. [Section 4.2, Eq. (11)] The asymptotic claim in Eq. (11) is mathematically incorrect. Let a = v_k ||ε − ε_θ(x_k,t_k)||_1 and d(a) = exp(−a) − 1 + a. As t_k → 0, for fixed ε−ε_θ, v_k → 0 and hence a → 0. In this limit d(a) ~ a^2/2, so a/d(a) ~ 2/a → ∞, not 1. Equation (11) therefore does not show that the Laplace loss behaves like a weighted L1 norm when t is small. Equation (12) is correct as ||r_k||_1 → ∞, but it does not rescue Eq. (11). This is not merely cosmetic: the interpretation of the non-Gaussian losses as L1-type objectives is part of the paper's motivation and is used to explain the Laplace models' behavior.
minor comments (5)
  1. [Abstract, Eq. (4), Definition 1] The sign convention for Δx_k is inconsistent: the abstract and Eq. (4) define Δx_k := x_k − x_{k+1}, while Definition 1 and the appendices use Δx_k := x_{k+1} − x_k. This creates confusion in the derivation of the reverse losses and should be fixed throughout.
  2. [Section 4.3, Eq. (13), Table 1] The constant C is reported inconsistently. Equation (13) states C = 1/2(1 + log(π/6)) ≈ 0.34, but 1/2(1 + log(π/6)) ≈ 0.18, while Table 1 writes 1/2(1 + log√(π/6)) ≈ 0.34. The numerical value and the expression should be reconciled.
  3. [Appendix A.4] The 'Linear Sub-Gaussian' KL derivation is incomplete and is not connected to any loss in Section 4 or Table 1; the section ends mid-derivation, and the notation s_1, s_2 is never defined in the main text. Either complete the derivation or remove the section.
  4. [Section 5, Table 2] The experimental support is difficult to assess. Table 2 reports BPD and FID for four configurations, but the footnote says these metrics are computed with different architectures, no standard DDPM/VDM baselines are given, no error bars or training details are provided, and the sentence in Section 5 mentions down-sampled ImageNet but no ImageNet results are shown.
  5. [Section 3.2] The paper states that 'we verify that we can recover Donsker's theorem when we let f = 0 and g = 1,' but no derivation is shown. Since Donsker's theorem requires independent increments, while the increments in Definition 1 are not independent when f ≠ 0, a brief verification or reference would be helpful.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Section 4 losses are closed-form KL divergences with no fitted constants, and Theorem 3.2 is an external convergence result; the unproven ELBO (5) is a soundness gap, not a circular step.

full rationale

The paper's derivation chain is self-contained with respect to its own inputs. The proposed losses in Section 4 (Eqs. 9, 10, 13, 14) are computed in closed form from the KL divergences between chosen increment distributions, with the derivations in Appendix A.5 and the supporting KL formulas in Appendices A.1-A.3; no parameter is fitted to the experimental results and then renamed as a prediction. Theorem 3.2 is an external weak-convergence claim whose proof rests on Billingsley (2013), Kloeden and Platen (1992), and standard central-limit and Slutsky arguments; it does not assume that the reverse conditionals of the non-Gaussian chain are Gaussian or that the specific Laplace/Uniform families of Section 4 are the correct variational posteriors, so those choices are not forced by the theorem. The paper contains no self-citations by the author, and the experiments compare against external benchmarks (CIFAR-10, down-sampled ImageNet) using standard architectures, rather than evaluating an in-sample fit. The most serious concern is not circularity but an unproven premise: Eq. (5) is simply asserted as a likelihood bound ('Letting q be the density of the Gaussian process Eq. 2, Eqs. (3) and (4) result in the same likelihood bound'), and Lemmas A.1-A.4 condition on x_k rather than the x_{k+1} appearing in Eq. (5). If Eq. (5) is not a valid ELBO for non-Gaussian increments, then the Section 4 losses are unjustified, but that is a soundness or correctness gap, not an instance of the derivation reducing to its own inputs by construction. No step defines X in terms of Y, fits a parameter and calls it a prediction, or imports a conclusion through self-citation. Accordingly, the circularity score is 0.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new fitted free parameters; the distribution family choice is a design decision, not a fitted quantity. The theoretical framework rests on standard regularity assumptions plus an unproved ELBO that is specific to this work. No new physical or mathematical entities are postulated.

assumptions (4)
  • domain assumption Assumption 1: f and g are Lipschitz in x and t
    Invoked in the tightness proof and Lemma B.5; standard for SDE discretization, cited from Saerkkae and Solin.
  • domain assumption Assumption 2: f and g satisfy linear growth bounds
    Used to bound moments of the increments in Eq. (68) and Lemma B.3.
  • domain assumption Assumption 3: E||z_k||^4 < K for the increment noise
    Needed for the fourth moment tightness criterion; excludes heavy tailed increments with infinite fourth moment.
  • ad hoc to paper ELBO Eq. (5) is valid for arbitrary increment distributions
    Stated without proof; it is the basis for all derived loss functions in Section 4, but is not the standard diffusion ELBO and is not justified.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Non-Normal Diffusion Models." pith.science (2026). https://pith.science/paper/LOUOBERG

@misc{pith2026241207935,
  author       = {Pith},
  title        = {Pith review of: Non-Normal Diffusion Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LOUOBERG}},
  note         = {Machine review of arXiv:2412.07935}
}
abstract

Diffusion models generate samples by incrementally reversing a process that turns data into noise. We show that when the step size goes to zero, the reversed process is invariant to the distribution of these increments. This reveals a previously unconsidered parameter in the design of diffusion models: the distribution of the diffusion step $\Delta x_k := x_{k} - x_{k + 1}$. This parameter is implicitly set by default to be normally distributed in most diffusion models. By lifting this assumption, we generalize the framework for designing diffusion models and establish an expanded class of diffusion processes with greater flexibility in the choice of loss function used during training. We demonstrate the effectiveness of these models on density estimation and generative modeling tasks on standard image datasets, and show that different choices of the distribution of $\Delta x_k$ result in qualitatively different generated samples.

Figures

Figures reproduced from arXiv: 2412.07935 by the authors.

Figure 1
Figure 1. Images generated from the same seed via (in order from top to bottom) Gaussian-Gaussian, Laplace-Laplace, Uniform￾Gaussian, and Uniform-Laplace diffusion increments. While the qualitative difference is somewhat subtle, Laplace diffusion ap￾pears to be biased towards smoother images with more saturated colors. 5. Experiments For illustrative purposes, we evaluate the diffusion models proposed in Section 4 on the CIFA… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 14 canonical work pages

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION format.date year duplicate empty "emp...

  2. [2]

    Anderson, B. D. Reverse-time diffusion equation models. Stochastic Processes and their Applications, 12 0 (3): 0 313--326, 1982

    1982

  3. [3]

    Convergence of probability measures

    Billingsley, P. Convergence of probability measures. John Wiley & Sons, 2013

    work page 2013

  4. [4]

    and Georges, A

    Bouchaud, J.-P. and Georges, A. Anomalous diffusion in disordered media: statistical mechanisms, models and physical applications. Physics reports, 195 0 (4-5): 0 127--293, 1990

    work page 1990

  5. [5]

    Large scale gan training for high fidelity natural image synthesis

    Brock, A., Donahue, J., and Simonyan, K. Large scale gan training for high fidelity natural image synthesis. arXiv preprint arXiv:1809.11096, 2018

    arXiv 2018

  6. [6]

    Very deep vaes generalize autoregressive models and can outperform them on images

    Child, R. Very deep vaes generalize autoregressive models and can outperform them on images. arXiv preprint arXiv:2011.10650, 2020

    arXiv 2011

  7. [7]

    Generating long sequences with sparse transformers

    Child, R., Gray, S., Radford, A., and Sutskever, I. Generating long sequences with sparse transformers. arXiv preprint arXiv:1904.10509, 2019

    arXiv 1904

  8. [8]

    Cugliandolo, L. F. Dynamics of glassy systems. arXiv preprint cond-mat/0210312, 2002

    arXiv 2002

Show all 39 references
  1. [9]

    and Nichol, A

    Dhariwal, P. and Nichol, A. Diffusion models beat gans on image synthesis. Advances in Neural Information Processing Systems, 34: 0 8780--8794, 2021

    2021

  2. [10]

    Nice: Non-linear independent components estimation

    Dinh, L., Krueger, D., and Bengio, Y. Nice: Non-linear independent components estimation. arXiv preprint arXiv:1410.8516, 2014

    2014 arXiv

  3. [11]

    Density estimation using real nvp

    Dinh, L., Sohl-Dickstein, J., and Bengio, S. Density estimation using real nvp. arXiv preprint arXiv:1605.08803, 2016

    2016 arXiv

  4. [12]

    Ethier, S. N. and Kurtz, T. G. Markov processes: characterization and convergence. John Wiley & Sons, 2009

    2009

  5. [13]

    Anomalous diffusion on percolating clusters

    Gefen, Y., Aharony, A., and Alexander, S. Anomalous diffusion on percolating clusters. Physical Review Letters, 50 0 (1): 0 77, 1983

    1983

  6. [14]

    Generative adversarial networks

    Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., and Bengio, Y. Generative adversarial networks. Communications of the ACM, 63 0 (11): 0 139--144, 2020

    2020

  7. [15]

    Gans trained by a two time-scale update rule converge to a local nash equilibrium

    Heusel, M., Ramsauer, H., Unterthiner, T., Nessler, B., and Hochreiter, S. Gans trained by a two time-scale update rule converge to a local nash equilibrium. In Guyon, I., Luxburg, U. V., Bengio, S., Wallach, H., Fergus, R., Vishwanathan, S., and Garnett, R. (eds.), Advances i...

    2017

  8. [16]

    and Barkai, E

    Hidalgo-Soria, M. and Barkai, E. Hitchhiker model for laplace diffusion processes. Physical Review E, 102 0 (1): 0 012109, 2020

    2020

  9. [17]

    Denoising diffusion probabilistic models

    Ho, J., Jain, A., and Abbeel, P. Denoising diffusion probabilistic models. Advances in Neural Information Processing Systems, 33: 0 6840--6851, 2020

    2020

  10. [18]

    J., Norouzi, M., and Salimans, T

    Ho, J., Saharia, C., Chan, W., Fleet, D. J., Norouzi, M., and Salimans, T. Cascaded diffusion models for high fidelity image generation. J. Mach. Learn. Res., 23 0 (47): 0 1--33, 2022

    2022

  11. [19]

    Consistency of pseudolikelihood estimation of fully visible boltzmann machines

    Hyv \"a rinen, A. Consistency of pseudolikelihood estimation of fully visible boltzmann machines. Neural Computation, 18 0 (10): 0 2283--2292, 2006

    2006

  12. [20]

    and Dayan, P

    Hyv \"a rinen, A. and Dayan, P. Estimation of non-normalized statistical models by score matching. Journal of Machine Learning Research, 6 0 (4), 2005

    2005

  13. [21]

    A style-based generator architecture for generative adversarial networks

    Karras, T., Laine, S., and Aila, T. A style-based generator architecture for generative adversarial networks. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp.\ 4401--4410, 2019

    2019

  14. [22]

    Elucidating the design space of diffusion-based generative models

    Karras, T., Aittala, M., Aila, T., and Laine, S. Elucidating the design space of diffusion-based generative models. arXiv preprint arXiv:2206.00364, 2022

    2022 arXiv

  15. [23]

    Variational diffusion models

    Kingma, D., Salimans, T., Poole, B., and Ho, J. Variational diffusion models. Advances in neural information processing systems, 34: 0 21696--21707, 2021

    2021

  16. [24]

    Kingma, D. P. and Dhariwal, P. Glow: Generative flow with invertible 1x1 convolutions. Advances in neural information processing systems, 31, 2018

    2018

  17. [25]

    Kingma, D. P. and Welling, M. Auto-encoding variational bayes. arXiv preprint arXiv:1312.6114, 2013

    2013 arXiv

  18. [26]

    Kloeden, P. E. and Platen, E. Stochastic differential equations. Springer, 1992

    1992

  19. [27]

    Learning multiple layers of features from tiny images

    Krizhevsky, A., Hinton, G., et al. Learning multiple layers of features from tiny images. Technical Report, 2009

    2009

  20. [28]

    and ksendal, B

    ksendal, B. and ksendal, B. Stochastic differential equations. Springer, 2003

    2003

  21. [29]

    Oord, A. v. d., Dieleman, S., Zen, H., Simonyan, K., Vinyals, O., Graves, A., Kalchbrenner, N., Senior, A., and Kavukcuoglu, K. Wavenet: A generative model for raw audio. arXiv preprint arXiv:1609.03499, 2016

    2016 arXiv

  22. [30]

    Salimans, T., Karpathy, A., Chen, X., and Kingma, D. P. Pixelcnn++: Improving the pixelcnn with discretized logistic mixture likelihood and other modifications. arXiv preprint arXiv:1701.05517, 2017

    2017 arXiv

  23. [31]

    and Solin, A

    S \"a rkk \"a , S. and Solin, A. Applied stochastic differential equations, volume 10. Cambridge University Press, 2019

    2019

  24. [32]

    Deep unsupervised learning using nonequilibrium thermodynamics

    Sohl-Dickstein, J., Weiss, E., Maheswaranathan, N., and Ganguli, S. Deep unsupervised learning using nonequilibrium thermodynamics. In International Conference on Machine Learning, pp.\ 2256--2265. PMLR, 2015

    2015

  25. [33]

    and Ermon, S

    Song, Y. and Ermon, S. Generative modeling by estimating gradients of the data distribution. Advances in Neural Information Processing Systems, 32, 2019

    2019

  26. [34]

    Sliced score matching: A scalable approach to density and score estimation

    Song, Y., Garg, S., Shi, J., and Ermon, S. Sliced score matching: A scalable approach to density and score estimation. In Uncertainty in Artificial Intelligence, pp.\ 574--584. PMLR, 2020 a

    2020

  27. [35]

    P., Kumar, A., Ermon, S., and Poole, B

    Song, Y., Sohl-Dickstein, J., Kingma, D. P., Kumar, A., Ermon, S., and Poole, B. Score-based generative modeling through stochastic differential equations. arXiv preprint arXiv:2011.13456, 2020 b

    2011 arXiv

  28. [36]

    Stroock, D. W. An introduction to Markov processes, volume 230. Springer Science & Business Media, 2013

    2013

  29. [37]

    and Kautz, J

    Vahdat, A. and Kautz, J. Nvae: A deep hierarchical variational autoencoder. Advances in neural information processing systems, 33: 0 19667--19679, 2020

    2020

  30. [38]

    Pixel recurrent neural networks

    Van Den Oord, A., Kalchbrenner, N., and Kavukcuoglu, K. Pixel recurrent neural networks. In International conference on machine learning, pp.\ 1747--1756. PMLR, 2016

    2016

  31. [39]

    and Romanovsky, M

    Vidov, P. and Romanovsky, M. Y. Analytical representations of non-gaussian laws of random walks. Physics of wave phenomena, 17: 0 218--227, 2009

    2009

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.