Pith. sign in

REVIEW 3 major objections 6 minor 45 references

Non-Reversible Langevin Algorithms for Constrained Sampling

T0 review · 3 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read Adding a non-reversible drift to a reflected Langevin diffusion on a convex domain preserves the constrained Gibbs target and provably accelerates convergence to it.

desk verdict The paper has a genuinely new idea for constrained sampling, but the central discrete algorithm is mis-specified: Eq. (1.9) defines a support map, not the skew projection the rest of the paper relies on. read the letter →

arxiv 2501.11743 v2 pith:WO7QXAFH submitted 2025-01-20 cs.LG math.PRstat.CO

Hengrong Du , Qi Feng , Changwei Tu , Xiaoyu Wang , Lingjiong Zhu This is my paper · ORCID
classification cs.LGmath.PRstat.CO MSC 60J6060H1065C05
keywords constrainedsamplingnon-reversibleLangevindynamicsskewreflectionprojectedMonteCarlo1-WassersteindistancetotalvariationspectralgapSkorokhodproblem
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

Constrained sampling asks for random draws from a target distribution $\pi(x)\propto e^{-f(x)}$ living on a bounded convex set. This paper proposes a way to speed that up: add a non-reversible (anti-symmetric) drift $J$ to the usual Langevin dynamics and reflect the process off the boundary along a skew direction tied to $J$. The authors prove the resulting continuous-time process still has $\pi$ as its stationary distribution, and that it approaches $\pi$ at an exponential rate at least as fast as the reversible reflected process, with strict acceleration exhibited in explicit quadratic examples. They also analyze the discrete algorithm and give a 1-Wasserstein convergence guarantee whose iteration complexity improves on projected Langevin Monte Carlo when reversibility is broken. If correct, this is a quantitative acceleration mechanism for constrained sampling in Bayesian and machine-learning settings.

What carries the argument

The load-bearing mechanism is the skew projection $P^J_C(x)=\operatorname{argmin}_{y\in\overline{C}}\langle y-x,\nu_J(P_C(x))\rangle$ with skew normal $\nu_J=(I+J)\nu/\sqrt{\lVert\nu\rVert^2+\lVert J\nu\rVert^2}$, combined with the anti-symmetric drift $J$ inside the domain. This projection returns updates to $C$ along a direction rotated by $J$, so the reflection term $\nu_J(X_t)L(dt)$ enforces the constraint while leaving the constrained Gibbs measure $\pi\propto e^{-f}$ invariant: in the generator $Lg=-\langle\nabla g,(I+J)\nabla f\rangle+\Delta g$ subject to $\nabla g\cdot\nu_J=0$, the $J$ contribution cancels by anti-symmetry and $\nabla\cdot J=0$. The proof of acceleration then runs through the spectral-gap comparison $\lambda_J\ge\lambda_0$ and, in the quadratic case, a synchronous coupling under the weighted norm $\lVert\cdot\rVert_{(I+J)^{-1}}$ that yields the rate $\lambda/C_{(I+J)^{-1}}$.

What would settle it

Take $C=[-1,1]^2$, a fixed nonzero anti-symmetric $2\times 2$ matrix $J$, and a point $x$ outside $C$ near a corner; compute the skew-projection $P^J_C$ defined in equation (1.9) and test whether $x-P^J_C(x)$ is parallel to $R^J_C(x)-P_C(x)$. A single counterexample would show that the skew-projection is not the oblique reflection that the SDE analysis requires.

Watch

Extended reading notes

Core claim

The central discovery is that non-reversibility and reflecting constraints are compatible. Define the skew normal $\nu_J=(I+J)\nu/\sqrt{\lVert \nu \rVert^2+\lVert J\nu \rVert^2}$ and the skew projection $P^J_C(x)=\operatorname{argmin}_{y\in\overline{C}}\langle y-x,\nu_J(P_C(x))\rangle$; then the SDE $dX_t=-(I+J(X_t))\nabla f(X_t)dt+\sqrt{2}\,dW_t+\nu_J(X_t)L(dt)$ admits $\pi\propto e^{-f}$ on $C$ as its invariant measure. The generator is $Lg=-\langle\nabla g,(I+J)\nabla f\rangle+\Delta g$ with Neumann boundary condition $\nabla g\cdot \nu_J=0$, and the added $J$ term contributes nothing to the invariant measure because $J$ is anti-symmetric and divergence-free. The paper proves exponential convergence in total variation and 1-Wasserstein distance with rate $\rho_J\ge\lambda_J\ge\lambda_0=\rho_0$, where $\lambda_0$ is the spectral gap of the reversible dynamics, and for quadratic $f$ obtains the explicit weighted-norm rate $\lambda/C_{(I+J)^{-1}}$ with $C_{(I+J)^{-1}}\le 1$. For the discretized algorithm SRNLMC it bounds the 1-Wasserstein error by three terms (continuous convergence, discretization error controlled by a local-time estimate, and a Girsanov change-of-measure error) and derives an iteration complexity of $\widetilde O(\log(1/\rho_J)d^{10}/(\rho_J^7\varepsilon^8))$, which is better than the reversible projected Langevin Monte Carlo baseline.

Load-bearing premise

The argument depends on the skew-projection $P^J_C$ acting as a genuine oblique projection onto $C$ along the skew normal, namely on the parallelity $x-P^J_C(x)\parallel R^J_C(x)-P_C(x)$; as written, $P^J_C$ is the argmin of a linear functional over $C$, which lands on an extreme boundary point and does not generally satisfy that relation, so the reflected dynamics and the convergence bounds built on that geometry are not well-defined if the relation fails.

Editorial extensions

If this is right

  • Non-reversibility can be introduced in constrained Langevin sampling without changing the target: the constrained Gibbs measure $\pi\propto e^{-f}$ remains invariant for SRNLD for any bounded anti-symmetric Lipschitz $J$.
  • Continuous-time SRNLD converges to $\pi$ in total variation and 1-Wasserstein distance at an exponential rate at least as fast as the reversible reflected Langevin diffusion, under only a spectral-gap assumption on the reversible dynamics rather than a log-concavity assumption.
  • For quadratic targets the convergence rate is explicit, $\lambda/C_{(I+J)^{-1}}$ with $C_{(I+J)^{-1}}\le 1$, so a nonzero anti-symmetric $J$ strictly accelerates mixing in that case.
  • SRNLMC reaches $\varepsilon$ accuracy in 1-Wasserstein distance in $\widetilde O(\log(1/\rho_J)d^{10}/(\rho_J^7\varepsilon^8))$ iterations, improving on the reversible PLMC complexity because $\rho_J\ge\rho_0$.
  • The discretization guarantee is made quantitative by a new bound on the boundary local time, $E[L(t)]\le C_L+C_L(1+d)t$, which controls the error introduced at reflections.

Reading between the lines

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

  • Inference: the achievable acceleration depends on the choice of $J$; choosing $J$ to maximize $\rho_J-\rho_0$ (or to minimize $C_{(I+J)^{-1}}$ in the quadratic case) is a design problem the paper leaves open, and the experiments tune the single parameter $a$ by hand.
  • Inference: the skew-projection construction should transfer to stochastic-gradient samplers; the paper tests the stochastic variant SRNSGLD empirically but gives no discretization theory for noisy gradients.
  • Inference: the method's benefit should be largest when the reversible spectral gap $\rho_0$ is small, because $J$ opens an additional drift channel for mixing; quantifying how $\rho_J-\rho_0$ grows with a given $J$ for nontrivial domains is a natural next step.
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.

Referee Report

3 major / 6 minor

Summary. The paper proposes skew-reflected non-reversible Langevin dynamics (SRNLD), a continuous-time diffusion on a bounded convex domain with a skew reflection at the boundary, and a discrete algorithm SRNLMC obtained by a 'skew projection' step. The main claims are that the Gibbs distribution constrained to the domain is invariant, that the continuous-time process converges exponentially in total variation and 1-Wasserstein distance, that breaking reversibility gives faster convergence than the reversible reflected Langevin dynamics, and that the discretized algorithm has explicit non-asymptotic bounds and better iteration complexity than projected Langevin Monte Carlo. Numerical experiments on a truncated Gaussian and on Bayesian linear and logistic regression are reported to illustrate the acceleration.

Significance. If the main claims were correct, the paper would make a genuinely useful contribution: it attempts to extend the known acceleration of non-reversible Langevin samplers to constrained sampling, and it provides non-asymptotic rates in both total variation and 1-Wasserstein metrics. The topic is timely and the ambition is appropriate. However, the central geometric construction used to define the discrete algorithm is not a projection with the claimed property, and the invariance proof silently depends on a divergence-free condition on J that is not among the stated assumptions. Because the discretization error analysis, the iteration complexity, and the headline comparison with PLMC are all built on that construction, the main results as stated do not apply to a well-defined algorithm. The paper does not provide code or machine-checked proofs, and the numerical Wasserstein comparisons are presented without confidence intervals. I therefore cannot recommend acceptance or even a minor revision; the central flaw is load-bearing.

major comments (3)
  1. [Eq. (1.9)] The map P^J_C defined by P^J_C(x) := argmin_{y in \bar C} <y - x, nu_J(P_C(x))> is not the oblique projection along nu_J that the paper needs. For a fixed x, the objective is linear in y, so the argmin is a support point of \bar C in the direction -nu_J(P_C(x)); it depends on P_C(x) only and not on the position of x along the normal ray. Consequently, the sentence after (1.9) asserting that x - P^J_C(x) is parallel to R^J_C(x) - P_C(x) is false in general. Example: let C be the unit disk in R^2, J = [[0,a],[-a,0]], and x = (2,0). Then P_C(x) = (1,0), and R^J_C(x) - P_C(x) = (I+J)(P_C(x)-x) = (-1,a). Equation (1.9) gives P^J_C(x) = argmin_{|y| <= 1} <y,(-1,a)> = (1,-a)/sqrt(1+a^2). For a = 1, x - P^J_C(x) = (2 - 1/sqrt(2), 1/sqrt(2)), which is not a scalar multiple of (-1,1). Thus the discrete update (1.10) does not implement the skew-reflected dynamics (1.5); it jumps to a boundary support point. As a result, the discretization estimates in Lemma 2.20, Corollary 2.22, Lemma 2.23, Theorem 2.24, and Corollary 2.25 analyze a different, unspecified map. This is load-bearing because SRNLMC is the practical algorithm of the paper and every guarantee about it depends on P^J_C having the claimed geometric meaning.
  2. [Theorem 2.6 and Remarks 2.8-2.9] The invariance proof uses the identity <grad, J^top grad g> = 0 and justifies it solely by antisymmetry of J. For a state-dependent matrix field J(x), this identity holds only if J is divergence-free in the sense that the row-wise divergence of J vanishes on each column. The paper first introduces the condition 'grad . J = 0' in Remarks 2.8 and 2.9, but it is not stated in Assumption 2.2 or anywhere in the standing assumptions. Without that condition, grad . (J^top grad g) = (div J) . grad g + sum_{i,j} J_{ji} partial_{ij} g, and the first term does not vanish. Thus Theorem 2.6, Lemma 2.7, and Equation (2.11) require an additional hypothesis on J. Since the continuous-time analysis and the definition of the spectral gap lambda_J rest on the invariant measure being pi, this is a load-bearing gap in the paper's main theorem.
  3. [Theorem 2.11 and Remark 2.26] The comparison with PLMC in Remark 2.26 uses only the inequality rho_J >= rho_0 from Theorem 2.11. This inequality is not strict for a general J, so the complexity bound in Corollary 2.25 does not imply that SRNLMC has a strictly lower iteration complexity than PLMC; it could be the same if rho_J = rho_0. Moreover, complexity is expressed through the unknown spectral gap rho_J, and the paper explicitly says after Theorem 2.11 that no explicit bound on rho_J is available. The quadratic example in Proposition 2.16 does give an explicit rate for constant H and J, but it does not cover the general non-convex setting of Theorem 2.24. Therefore the assertion that breaking reversibility yields better convergence guarantees than the reversible PLMC is not established by the stated results.
minor comments (6)
  1. [Section 2.1, Lemma 2.3 proof] There is a typo 'the soluiton is unique'; also the proof jumps from the definition of nu_J^s to existence without verifying the conditions of the cited Tanaka/Lions-Sznitman theorems in the text.
  2. [Equation (2.29)] The expression '(Xt - eX ^top_t )' contains a typo and should be '(Xt - eXt)^top'.
  3. [Lemma 2.19 and Corollary 2.21] The notation for the log factors is inconsistent: Lemma 2.19 has (log(T/eta))^{1/2}, while Lemma 2.20 and Corollary 2.21 use (log(T/eta))^{1/4}; please state the exact intermediate inequalities.
  4. [Section 3.1] The 1-Wasserstein distances are computed separately for each dimension, not in the full R^3 metric; the text should say so explicitly because this weakens the visual comparison.
  5. [Numerical experiments] The Wasserstein convergence plots and accuracy plots are shown without error bars or multiple seeds; please report standard errors or confidence intervals so the empirical acceleration can be assessed.
  6. [Lemma 2.13] The regularity argument invokes [Lie90] for oblique parabolic equations, but the boundary condition (2.3) involves the state-dependent skew normal nu_J; please justify that the cited result applies to this nonsymmetric, state-dependent boundary condition.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: bounds derive from stated assumptions; the main defect is a geometric misidentification, not a circular reduction.

full rationale

The derivation chain is self-contained against the stated assumptions. The continuous-time bound (Theorem 2.11) is obtained by defining rho_J as the TV decay rate and then proving rho_J >= lambda_J >= lambda_0 = rho_0 through the Dirichlet-form comparison and parabolic regularity; this is a substantive argument rather than a restatement of inputs. The discrete-time bounds (Corollary 2.22, Lemma 2.23, Theorem 2.24, Corollary 2.25) combine the continuous-time estimate with local-time and Skorokhod estimates; no quantity is fitted to data and no target value is used as an input. The comparison with PLMC is conditional on the unquantified spectral gap rho_J, which weakens the advertised acceleration, but rho_J is not a fitted parameter and the inequality rho_J >= rho_0 is proven, so the comparison is not circular. Cited external results ([Tan79], [LS84], [DI93], [Lie90], [BGT04], [HHMS05]) supply the analytic machinery; the authors' own previous work appears only as related literature and is not load-bearing. A genuine defect exists at Eq. (1.9): defining P^J_C as an argmin of a linear functional makes it a support-face map whose value depends only on P_C(x), so the asserted parallelism 'x - P^J_C(x) parallel to R^J_C(x) - P_C(x)' is false in general; consequently the discrete map (1.10) is not evidently the skew projection used in the continuous analysis. This is a correctness and well-posedness issue, not a circular reduction, and therefore does not raise the circularity score under the stated rules.

Assumptions & free parameters 2 free parameters · 6 assumptions · 1 invented entities

The central results rest on the bounded convex domain, Lipschitz regularity, a positive spectral gap, a uniform obliqueness condition, and an unstated divergence-free condition on J. The rates are expressed through unknown spectral gaps ρJ and λ0, so the theory does not provide readily computable guarantees. No physical entities beyond the skew reflection construction are introduced.

free parameters (2)
  • Anti-symmetric matrix field J(x) = a = 1, 1.5, 2 in experiments for constant J_a
    The user chooses the skew matrix J; the convergence rate ρJ, the constants C∞ and CL, and the skew projection all depend on it. No optimal choice is derived for constrained domains, and experiments hand-tune a scalar a.
  • Stepsize η = 10^{-4} in all experiments
    Standard algorithm parameter chosen by hand. The complexity bounds in Corollary 2.25 depend on η, but it is not fit to data to force the theory.
assumptions (6)
  • domain assumption λ0 > 0, the spectral gap of the reversible reflected Langevin generator on C
    Assumption 2.10; Theorem 2.11 and Corollary 2.25 express all rates through ρJ and λ0, so without this the convergence bounds are vacuous.
  • domain assumption C is bounded, convex, has C1 boundary, contains balls of radius r and R centered at 0
    Assumption 2.1; used for boundedness, the Skorokhod problem, the W1 ≤ 2R TV inequality, and the gauge norm estimates.
  • domain assumption ∇f and J are Lipschitz with constants L and LJ
    Assumption 2.2; needed for strong well-posedness, the drift bounds, and the discretization error estimates.
  • domain assumption The skew reflection satisfies ⟨νJ, ν⟩ ≥ δ0 > 0
    Stated after Eq. (1.6); used in Lemma 2.17 to bound the boundary local time, and implicitly for well-posedness of the oblique reflection.
  • ad hoc to paper ∇·J = 0, the matrix field J is divergence-free
    Invoked in Remark 2.8 and the proof of Theorem 2.6 to establish Gibbs invariance and the Dirichlet form identity. Not stated in Assumption 2.2, so for state-dependent J it is a hidden assumption that can break the central invariant-measure claim.
  • standard math Regularity estimate for oblique parabolic equations (Lie90) gives Hölder continuity of the density ratio p1(x,·)
    Used in Lemma 2.13 to ensure p1(x,·) ∈ L2(π), which is needed for the TV convergence argument.
invented entities (1)
  • Skew unit normal vector νJ
    purpose: Defines the oblique reflection direction that keeps the process inside C while preserving the Gibbs invariant measure.
    A mathematical construction introduced by the paper, not an empirically observable entity. Its validity depends on the boundedness of J, the angle condition, and the unstated divergence-free condition for state-dependent J.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Non-Reversible Langevin Algorithms for Constrained Sampling." pith.science (2026). https://pith.science/paper/WO7QXAFH

@misc{pith2026250111743,
  author       = {Pith},
  title        = {Pith review of: Non-Reversible Langevin Algorithms for Constrained Sampling},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WO7QXAFH}},
  note         = {Machine review of arXiv:2501.11743}
}
read the original abstract

We consider the constrained sampling problem where the goal is to sample from a target distribution on a constrained domain. We propose skew-reflected non-reversible Langevin dynamics (SRNLD), a continuous-time stochastic differential equation with skew-reflected boundary. We obtain non-asymptotic convergence rate of SRNLD to the target distribution in both total variation and 1-Wasserstein distances. By breaking reversibility, we show that the convergence is faster than the special case of the reversible dynamics. Based on the discretization of SRNLD, we propose skew-reflected non-reversible Langevin Monte Carlo (SRNLMC), and obtain non-asymptotic discretization error from SRNLD, and convergence guarantees to the target distribution in 1-Wasserstein distance. We show better performance guarantees than the projected Langevin Monte Carlo in the literature that is based on the reversible dynamics. Numerical experiments are provided for both synthetic and real datasets to show efficiency of the proposed algorithms.

Figures

Figures reproduced from arXiv: 2501.11743 by the authors.

Figure 1
Figure 1. Visualized density plots for the first 2 dimensions in ball constraint [PITH_FULL_IMAGE:figures/full_fig_p024_1.png] view at source ↗
Figure 2
Figure 2. 1-Wasserstein distance in each dimension of PLMC and SRNLMC in ball constraint 24 [PITH_FULL_IMAGE:figures/full_fig_p024_2.png] view at source ↗
Figure 3
Figure 3. Visualized density plots for the first 2 dimensions in cubic constraint [PITH_FULL_IMAGE:figures/full_fig_p025_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: 1-Wasserstein distance in each dimension of PLMC and SRNLMC in cubic constraint We can observe from [PITH_FULL_IMAGE:figures/full_fig_p025_4.png]
Figure 5
Figure 5. Figure 5: Prior and posterior distributions plot with disk constraint [PITH_FULL_IMAGE:figures/full_fig_p027_5.png]
Figure 6
Figure 6. Figure 6: MSE result of SRNSGLD and PSGLD for the constrained Bayesian linear regression [PITH_FULL_IMAGE:figures/full_fig_p027_6.png]
Figure 7
Figure 7. Figure 7: Accuracy over the training set and the test set for the synthetic data. The red part denotes [PITH_FULL_IMAGE:figures/full_fig_p029_7.png]
Figure 8
Figure 8. Figure 8: Accuracy over the training set and the test set for the telescope dataset. The red part [PITH_FULL_IMAGE:figures/full_fig_p030_8.png]
Figure 9
Figure 9. Figure 9: Accuracy over the training set and the test set for the Titanic dataset. The red part denotes [PITH_FULL_IMAGE:figures/full_fig_p030_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

45 extracted references · 39 canonical work pages

  1. [1]

    Efficient constrained sampling via the mirror- L angevin algorithm

    Kwangjun Ahn and Sinho Chewi. Efficient constrained sampling via the mirror- L angevin algorithm. In Advances in Neural Information Processing Systems (NeurIPS) , volume 34, 2021

    work page 2021

  2. [2]

    An introduction to MCMC for machine learning

    Christophe Andrieu, Nando De Freitas, Arnaud Doucet, and Michael I Jordan. An introduction to MCMC for machine learning. Machine Learning , 50(1):5--43, 2003

    2003

  3. [3]

    A simple proof of the P oincar \'e inequality for a large class of probability measures

    Dominique Bakry, Franck Barthe, Patrick Cattiaux, and Arnaud Guillin. A simple proof of the P oincar \'e inequality for a large class of probability measures. Electronic Communications in Probability , 13:60--66, 2008

    work page 2008

  4. [4]

    Sampling from a log-concave distribution with compact support with proximal L angevin M onte C arlo

    Nicolas Brosse, Alain Durmus, \' E ric Moulines, and Marcelo Pereyra. Sampling from a log-concave distribution with compact support with proximal L angevin M onte C arlo. In Proceedings of the 2017 Conference on Learning Theory , volume 65, pages 319--342. PMLR, 2017

    work page 2017

  5. [5]

    Finite-time analysis of projected L angevin M onte C arlo

    Sebastien Bubeck, Ronen Eldan, and Joseph Lehec. Finite-time analysis of projected L angevin M onte C arlo. In Advances in Neural Information Processing Systems , volume 28, 2015

    work page 2015

  6. [6]

    Sampling from a log-concave distribution with projected L angevin M onte C arlo

    S \'e bastien Bubeck, Ronen Eldan, and Joseph Lehec. Sampling from a log-concave distribution with projected L angevin M onte C arlo. Discrete & Computational Geometry , 59(4):757--783, 2018

    work page 2018

  7. [7]

    Analysis and Geometry of Markov Diffusion Operators , volume 103

    Dominique Bakry, Ivan Gentil, and Michel Ledoux. Analysis and Geometry of Markov Diffusion Operators , volume 103. Springer, 2014

    work page 2014

  8. [8]

    A symmetrized E uler scheme for an efficient approximation of reflected diffusions

    Mireille Bossy, Emmanuel Gobet, and Denis Talay. A symmetrized E uler scheme for an efficient approximation of reflected diffusions. Journal of Applied Probability , 41(3):877--889, 2004

    work page 2004

Show all 45 references
  1. [9]

    A note on T alagrand’s transportation inequality and logarithmic S obolev inequality

    Patrick Cattiaux, Arnaud Guillin, and Li-Ming Wu. A note on T alagrand’s transportation inequality and logarithmic S obolev inequality. Probability Theory and Related Fields , 148:285--304, 2010

    2010

  2. [10]

    Exponential ergodicity of mirror- L angevin diffusions

    Sinho Chewi, Thibaut Le Gouic, Cheng Lu, Tyler Maunu, Philippe Rigollet, and Austin Stromme. Exponential ergodicity of mirror- L angevin diffusions. In Advances in Neural Information Processing Systems (NeurIPS) , volume 33, 2020

    2020

  3. [11]

    The S korohod oblique reflection problem in domains with corners and application to stochastic differential equations

    Cristina Costantini. The S korohod oblique reflection problem in domains with corners and application to stochastic differential equations. Probability Theory and Related Fields , 91:43--70, 1992

    1992

  4. [12]

    SDE s with oblique reflection on nonsmooth domains

    Paul Dupuis and Hitoshi Ishii. SDE s with oblique reflection on nonsmooth domains. Annals of Probability , 21(1):554--580, 1993

    1993

  5. [13]

    Duncan, Tony Leli\` e vre, and Grigoris A

    Andrew B. Duncan, Tony Leli\` e vre, and Grigoris A. Pavliotis. Variance reduction using nonreversible L angevin samplers. Journal of Statistical Physics , 163(3):457--491, 2016

    2016

  6. [14]

    Duncan, Grigoris A

    Andrew B. Duncan, Grigoris A. Pavliotis, and Konstantinos C. Zygalakis. Nonreversible L angevin samplers: Splitting schemes, analysis and implementation. arXiv preprint arXiv:1701.04247 , 2017

    2017 arXiv

  7. [15]

    Accelerating the diffusion-based ensemble sampling by non-reversible dynamics

    Futoshi Futami, Iseei Sato, and Masashi Sugiyama. Accelerating the diffusion-based ensemble sampling by non-reversible dynamics. In Proceedings of the 37th International Conference on Machine Learning , volume 119, pages 3337--3347. PMLR, 2020

    2020

  8. [16]

    Bayesian Data Analysis

    Andrew Gelman, John B Carlin, Hal S Stern, and Donald B Rubin. Bayesian Data Analysis . Chapman & Hall/CRC Press, 1995

    1995

  9. [17]

    Decentralized stochastic gradient L angevin dynamics and H amiltonian M onte C arlo

    Mert G\"urb\"uzbalaban, Xuefeng Gao, Yunhan Hu, and Lingjiong Zhu. Decentralized stochastic gradient L angevin dynamics and H amiltonian M onte C arlo. Journal of Machine Learning Research , 22(239):1--69, 2021

    2021

  10. [18]

    Breaking reversibility accelerates L angevin dynamics for global non-convex optimization

    Xuefeng Gao, Mert G\" u rb\" u zbalaban, and Lingjiong Zhu. Breaking reversibility accelerates L angevin dynamics for global non-convex optimization. In Advances in Neural Information Processing Systems (NeurIPS) , 2020

    2020

  11. [19]

    Penalized overdamped and underdamped L angevin M onte C arlo algorithms for constrained sampling

    Mert G\" u rb\" u zbalaban, Yuanhan Hu, and Lingjiong Zhu. Penalized overdamped and underdamped L angevin M onte C arlo algorithms for constrained sampling. Journal of Machine Learning Research , 25(263):1--67, 2024

    2024

  12. [20]

    Generalized EXTRA stochastic gradient L angevin dynamics

    Mert G\" u rb\" u zbalaban, Mohammad Rafiqul Islam, Xiaoyu Wang, and Lingjiong Zhu. Generalized EXTRA stochastic gradient L angevin dynamics. arXiv preprint arXiv:2412.01993 , 2024

    2024

  13. [21]

    Gibbs and Francis Edward Su

    Alison L. Gibbs and Francis Edward Su. On choosing and bounding probability metrics. International Statistical Review , 70(3):419--435, 2002

    2002

  14. [22]

    Accelerating G aussian diffusions

    Chii-Ruey Hwang, Shu-Yin Hwang-Ma, and Shuenn-Jyi Sheu. Accelerating G aussian diffusions. Annals of Applied Probability , 3:897--913, 1993

    1993

  15. [23]

    Accelerating diffusions

    Chii-Ruey Hwang, Shu-Yin Hwang-Ma, and Shuenn-Jyi Sheu. Accelerating diffusions. Annals of Applied Probability , 15:1433--1444, 2005

    2005

  16. [24]

    Mirrored L angevin dynamics

    Ya-Ping Hsieh, Ali Kavis, Paul Rolland, and Volkan Cevher. Mirrored L angevin dynamics. In Advances in Neural Information Processing Systems , volume 31, 2018

    2018

  17. [25]

    Non-convex stochastic optimization via non-reversible stochastic gradient L angevin dynamics

    Yuanhan Hu, Xiaoyu Wang, Xuefeng Gao, G\" u rb\" u zbalaban, and Lingjiong Zhu. Non-convex stochastic optimization via non-reversible stochastic gradient L angevin dynamics. arXiv:2004.02823 , 2020. 45 pages

    2004 arXiv

  18. [26]

    Projected stochastic gradient L angevin algorithms for constrained sampling and non-convex learning

    Andrew Lamperski. Projected stochastic gradient L angevin algorithms for constrained sampling and non-convex learning. In Conference on Learning Theory , volume 134, pages 2891--2937. PMLR, 2021

    2021

  19. [27]

    On the H \"o lder gradient estimate for solutions of nonlinear elliptic and parabolic oblique boundary value problems

    Gary A Lieberman. On the H \"o lder gradient estimate for solutions of nonlinear elliptic and parabolic oblique boundary value problems. Communications in Partial Differential Equations , 15(4):515--523, 1990

    1990

  20. [28]

    Optimal non-reversible linear drift for the convergence to equilibrium of a diffusion

    Tony Leli \`e vre, Francis Nier, and Grigorios A Pavliotis. Optimal non-reversible linear drift for the convergence to equilibrium of a diffusion. Journal of Statistical Physics , 152(2):237--274, Jul 2013

    2013

  21. [29]

    Stochastic differential equations with reflecting boundary conditions

    Pierre-Louis Lions and Alain-Sol Sznitman. Stochastic differential equations with reflecting boundary conditions. Communications on Pure and Applied Mathematics , 37(4):511--537, 1984

    1984

  22. [30]

    Vempala, and Andre Wibisono

    Ruilin Li, Molei Tao, Santosh S. Vempala, and Andre Wibisono. The mirror L angevin algorithm converges with vanishing bias. In Sanjoy Dasgupta and Nika Haghtalab, editors, Proceedings of The 33rd International Conference on Algorithmic Learning Theory , volume 167, pages 718--...

    2022

  23. [31]

    Irreversible L angevin samplers and variance reduction: a large deviation approach

    Luc Rey-Bellet and Konstantinos Spiliopoulos. Irreversible L angevin samplers and variance reduction: a large deviation approach. Nonlinearity , 28:2081, 2015

    2015

  24. [32]

    Variance reduction for irreversible L angevin samplers and diffusion on graphs

    Luc Rey-Bellet and Konstantinos Spiliopoulos. Variance reduction for irreversible L angevin samplers and diffusion on graphs. Electronic Communications in Probability , 20(15):16 pp., 2015

    2015

  25. [33]

    Improving the convergence of reversible samplers

    Luc Rey-Bellet and Konstantinos Spiliopoulos. Improving the convergence of reversible samplers. Journal of Statistical Physics , 164(3):472--494, 2016

    2016

  26. [34]

    Weak P oincar\' e inequalities and L^ 2 -convergence rates of M arkov semigroups

    Michael R\" o ckner and Feng-Yu Wang. Weak P oincar\' e inequalities and L^ 2 -convergence rates of M arkov semigroups. Journal of Functional Analysis , 185:546--603, 2001

    2001

  27. [35]

    Primal dual interpretation of the proximal stochastic gradient L angevin algorithm

    Adil Salim and Peter Richt\' a rik. Primal dual interpretation of the proximal stochastic gradient L angevin algorithm. In Advances in Neural Information Processing Systems (NeurIPS) , volume 33, 2020

    2020

  28. [36]

    ( N onsymmetric) D irichlet operators on L ^1 : existence, uniqueness and associated M arkov processes

    Wilhelm Stannat. ( N onsymmetric) D irichlet operators on L ^1 : existence, uniqueness and associated M arkov processes. Annali della Scuola Normale Superiore di Pisa-Classe di Scienze , 28(1):99--140, 1999

    1999

  29. [37]

    Inverse problems: A B ayesian perspective

    Andrew M Stuart. Inverse problems: A B ayesian perspective. Acta Numerica , 19:451--559, 2010

    2010

  30. [38]

    Stochastic differential equations with reflecting boundary condition in convex regions

    Hiroshi Tanaka. Stochastic differential equations with reflecting boundary condition in convex regions. Hiroshima Mathematical Journal , 9:163--177, 1979

    1979

  31. [39]

    Operator reverse monotonicity of the inverse

    Alexis Akira Toda. Operator reverse monotonicity of the inverse. The American Mathematical Monthly , 118(1):82--83, 2011

    2011

  32. [40]

    Consistency and fluctuations for stochastic gradient L angevin dynamics

    Yee Whye Teh, Alexandre H Thiery, and Sebastian J Vollmer. Consistency and fluctuations for stochastic gradient L angevin dynamics. The Journal of Machine Learning Research , 17(1):193--225, 2016

    2016

  33. [41]

    On estimation of the logarithmic S obolev constant and gradient estimates of heat semigroups

    Feng-Yu Wang. On estimation of the logarithmic S obolev constant and gradient estimates of heat semigroups. Probability Theory and Related Fields , 108:87--101, 1997

    1997

  34. [42]

    Sheng-Jhih Wu, Chii-Ruey Hwang, and Moody T. Chu. Attaining the optimal G aussian diffusion acceleration. Journal of Statistical Physics , 155(3):571--590, 2014

    2014

  35. [43]

    Constrained exploration via reflected replica exchange stochastic gradient L angevin dynamics

    Haoyang Zheng, Hengrong Du, Qi Feng, Wei Deng, and Guang Lin. Constrained exploration via reflected replica exchange stochastic gradient L angevin dynamics. In Proceedings of the 41st International Conference on Machine Learning , volume 235, pages 61321--61348. PMLR, 2024

    2024

  36. [44]

    Constrained L angevin algorithms with L -mixing external random variables

    Yuping Zheng and Andrew Lamperski. Constrained L angevin algorithms with L -mixing external random variables. In Advances in Neural Information Processing Systems (NeurIPS) , volume 35, 2022

    2022

  37. [45]

    Wasserstein control of mirror L angevin M onte C arlo

    Kelvin Shuangjian Zhang, Gabriel Peyr\' e , Jalal Fadili, and Marcelo Pereyra. Wasserstein control of mirror L angevin M onte C arlo. In Conference on Learning Theory , volume 125, pages 3814--3841. PMLR, 2020

    2020

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.