betaFeedback

WorksComputer Science

Numerical Integration on Bounded Domains via Learned Reflected Diffusions

Abstract

We study high-dimensional integration on bounded convex domains. Given a non-negative integrand rr, the goal is to estimate Z=∫Dr(x) dxZ=\int_{D}r(x)\,\mathrm dx on D⊂RdD\subset\mathbb R^d. Standard Quasi-Monte Carlo does not adapt its evaluation points to the integrand, while effective importance sampling requires a proposal that captures its concentration. Learned diffusion samplers can supply such a proposal, but existing methods are formulated on Rd\mathbb R^d and do not respect the domain constraint x∈Dx\in D. We instead build the proposal from reflected dynamics that stay in DD by construction. A reflected diffusion progressively transforms the target distribution toward the uniform distribution on DD, and a neural score model is trained to reverse this process. A change-of-measure formula for reflected diffusions yields a training objective without additional boundary terms and an unbiased importance-sampling estimator of ZZ. For non-smooth integrands, where reparameterization gradients may be undefined or uninformative, we derive a REINFORCE estimator of the training gradient using Gaussian proposal densities before reflection, without evaluating reflected transition densities or differentiating rr. We evaluate on numerous synthetic targets and real-world problems, demonstrating strong performance of the proposed method, as well as favorable scaling with the dimension of the problem and the number of points used for the Monte Carlo estimate.

Introduction

We consider the problem of estimating high-dimensional integrals over bounded constrained domains. Let D⊂RdD \subset\mathbb{R}^d be a bounded convex domain, and let r:D→R≥0r : D \to\mathbb{R}_{\ge0} be measurable and satisfy ∫Dr(x) dx<∞\int_D r(x)\,\mathrm{d}x < \infty. Our goal is to estimate the integral

Z=∫Dr(x) dx.(1)Z = \int_D r(x)\,\mathrm{d}x. \tag*{(1)}

Integrals of this form are a computational bottleneck across the sciences. In statistical physics and chemistry, free energies and thermodynamic observables are logarithms of such constants, and learned samplers have been used to estimate them for many-body systems and lattice field theories [58, 39] (Noé et al. 2019). In computer graphics, every pixel of a rendered image is an integral over light paths, parameterized so that the integration domain is the unit cube [38]. In computational finance, the price of a path-dependent derivative is an expectation over a discretized asset path, which the standard inverse-CDF construction turns into an integral over [0,1]d[0,1]^d with dd the number of monitoring dates [19 Dick et al.,] 2013). In Bayesian statistics, the model evidence is an integral of an unnormalized posterior [27, 26], confined to a bounded region whenever the parameters are subject to simplex constraints or under compactly supported priors.

These tasks give rise to integration problems on bounded domains, where the location of integrand mass is unknown in advance. This motivates samplers that respect the domain constraints and adapt to where the mass lies, including when it concentrates near faces or corners, or when the integrand vanishes on parts of the domain.

Existing estimators differ in how much they adapt to r(x)r(x). Standard Quasi-Monte Carlo [40, 14 QMC] spreads a deterministic point set over the cube without reference to the integrand. Bayesian quadrature [42, 46] adapts by placing a Gaussian process prior on r(x)r(x), but at a cost cubic in the number of evaluations [8]. Importance sampling relies on a proposal distribution whose choice determines the estimator’s variance [22]. Neural proposals learn that proposal from evaluations of r(x)r(x) alone, either with normalizing flows [47, 38] or with diffusion samplers [59, 52] (Berner et al. 2024). The latter avoids the invertibility constraints that limit flow architectures [28]. Diffusion samplers, however, are formulated on Rd\mathbb{R}^{d} with a Gaussian reference process throughout the existing work and do not respect the constraint x∈Dx \in\mathcal{D}. A more thorough discussion of prior work on numerical integration methods is presented in Appendix § 7.

To design a learnable diffusion-based proposal that respects the domain constraint, we draw inspiration from diffusion models [50] for constrained generative modeling. Existing approaches fall into two groups. The first changes the geometry in which the process evolves: mirror diffusion models map D\mathcal{D} onto Rd\mathbb{R}^{d} through the gradient of a strictly convex potential and run a standard diffusion in the resulting dual space [33], while log-barrier diffusions equip the interior of D\mathcal{D} with the Riemannian metric induced by the Hessian of a barrier function, under which the process remains in the interior [16]. The second group keeps the Euclidean geometry of D\mathcal{D} and confines trajectories by reflecting them at the boundary, using reflected Brownian motion as the noising process [16, 35]. In this work, we follow the second route and adapt reflected diffusions from generative modeling to sampling, where no samples from π\pi are available and the reverse dynamics must be learned from evaluations of rr alone.

We outline our contributions as follows:

  • We propose reflected diffusion samplers (RDS) for integration on bounded domains, extending control-based diffusion samplers [59, 52, 4] to constrained D\mathcal{D}. Our construction uses reflected Brownian motion with a time-dependent noise schedule. When initialized from the uniform distribution ν\nu on D\mathcal{D}, this process preserves uniform marginals [34] and defines a tractable reference path measure for the integral (1). Girsanov’s theorem [31] then transfers this representation to equivalent controlled reflected diffusions through path-space importance weights.

We show that the unique zero-variance ideal proposal corresponds to the same reflected noising dynamics initialized from the normalized target π(x)=r(x)/Z\pi(x)=r(x)/Z, and propose to learn its reversal with a neural network-parameterized drift. The learned reversal by construction also acts as the sampler from π\pi.

  • We design a training objective for the learned reversal to be the KL divergence to the ideal proposal. For smooth targets, its gradients can be computed via the reparameterization trick. It includes a running control cost with a terminal potential and no boundary term, closely resembling the optimal control objective of Zhang & Chen [59]. We also derive a REINFORCE gradient estimator with a leave-one-out baseline, which uses only the pre-reflection Gaussian transition densities and does not require gradients of r(x)r(x), applying to targets that are non-smooth or vanish to zero on part of D\mathcal{D}, where reparameterization gradients are uninformative. Integrands that vanish on part of the domain are handled by adding a positive offset and correcting the resulting integral estimate.

  • We provide rigorous theoretical derivation of all components of the proposed methodology, adapting existing theoretical results to the considered setting of the variance exploding reflected diffusion samplers. In addition, we prove a total-variation accuracy bound for the designed estimator, which is a reflected version of the bound presented in Guo et al. [21] for diffusion samplers in Rd\mathbb{R}^{d}.

  • We evaluate on Genz integrands [18], bounded synthetic targets, Feynman parameter integral [6] and Asian option pricing [36] against Quasi-Monte Carlo, Bayesian quadrature and normalizing flow importance sampling baselines. The experiments demonstrate strong performance of the proposed method, as well as favorable scaling with the dimension of the problem and the number of points used for the estimate. In addition, we compare against Path Integral Sampler (PIS, Zhang & Chen [59]) proposal by mapping Rd\mathbb{R}^{d} onto D‾\overline{D}, and demonstrate that the learned RDS proposal achieves lower integration errors against its unconstrained counterpart.

Source code: github.com/kkorolev1/reflected-diffusion-samplers.

Background on Reflected SDE

Denote D‾\overline{D} and ∂D\partial D to be the closure and boundary of DD, and vol⁡(D)\operatorname{vol}(D) to be its volume. Consider a reflected SDE on DD with drift f:D‾×[0,1]→Rdf:\overline{D}\times[0,1]\to\mathbb{R}^{d} and scalar diffusion coefficient g:[0,1]→Rg:[0,1]\to\mathbb{R} [45]:

dXt=f(Xt,t) dt+g(t) dWt+dLt.(2)\mathrm{d}X_t=f(X_t,t)\,\mathrm{d}t+g(t)\,\mathrm{d}W_t+\mathrm{d}L_t. \tag*{(2)}

Here WtW_t is a standard dd-dimensional Brownian motion. Under the smooth-boundary assumptions used for the theoretical results, a solution of SDE (2) has the form (X,ℓ)(X,\ell), where ℓ\ell is continuous, nondecreasing, starts from zero, and increases only when Xt∈∂DX_t\in\partial D. Throughout the manuscript, the associated vector reflection term is the derived process Lt:=∫0tnin(Xs) dℓsL_t:=\int_0^t\mathbf{n}_{\mathrm{in}}(X_s)\,\mathrm{d}\ell_s, where nin\mathbf{n}_{\mathrm{in}} is the inward unit normal. Thus dLt\mathrm{d}L_t pushes XX inward at boundary times.

Let μt:=Law⁡(Xt)\mu_t:=\operatorname{Law}(X_t) be the marginal law of XtX_t. Suppose that, for t>0t>0, μt\mu_t admits a density ptp_t and that these densities have the time, space, and boundary-trace regularity needed for the following identities to hold. Then ptp_t evolves according to the Fokker–Planck equation with a Neumann boundary condition [48 Section 2.6.2–2.6.3]:

∂tpt(x)=div⁡(−pt(x)f(x,t)+g2(t)2∇xpt(x)),x∈D, t∈(0,1],n(x)T(pt(x)f(x,t)−g2(t)2∇xpt(x))=0,x∈∂D, t∈(0,1],(3)\begin{aligned} \partial_t p_t(x)&=\operatorname{div}\left(-p_t(x)f(x,t)+\frac{g^2(t)}{2}\nabla_xp_t(x)\right), &&x\in D,\ t\in(0,1],\\ \mathbf{n}(x)^{\mathsf{T}}\left(p_t(x)f(x,t)-\frac{g^2(t)}{2}\nabla_xp_t(x)\right)&=0, &&x\in\partial D,\ t\in(0,1], \tag*{(3)} \end{aligned}

where n(x)\mathbf{n}(x) is the outward unit normal to ∂D\partial D. The boundary condition prevents probability mass from crossing the boundary, so the density evolves entirely within DD. The initial law is imposed separately through μt⇒μ0\mu_t\Rightarrow\mu_0 as t↓0t\downarrow0, which follows from path continuity. Even when μ0\mu_0 has a density p0p_0, this does not make (3) valid at t=0t=0 without additional initial regularity and boundary compatibility. Without the stated density and regularity, the forward equation is instead interpreted in its weak measure-valued form.

Reflected Diffusion Samplers

In this section, we present the main theoretical results and the derivation of the proposed method. Each statement points to a fully quantified statement and proof in the Appendix. The two distributions used in this section are the uniform distribution and the normalized target:

ν(x):=vol⁡(D)−1,π(x):=r(x)Z,x∈D‾.(4)\nu(x):=\operatorname{vol}(D)^{-1},\qquad\pi(x):=\frac{r(x)}{Z},\qquad x\in\overline{D}. \tag*{(4)}

Path-measure and training theoretical results assume measurable rr bounded above and away from zero. The full-interval score representation requires additional regularity. We later discuss zero- and negative-valued targets. See Appendix § 8.§ 1 for notation and conventions.

Reference process of the reflected diffusion

Our goal is to construct an importance-sampling estimator for (1) using stochastic dynamics that remain inside the constrained domain DD. To this end, we first introduce a simple reference process whose path measure will serve as a base measure with respect to which a proposal is defined. Ideally, this reference process should satisfy two properties. First, it should remain inside D‾\overline{D} by construction. Second, its marginal distribution should be known and easy to sample from. We obtain both properties with the reflected variance-exploding1 SDEs [35 RVE SDEs]. Whenever the SDE below is well posed, each initial law determines a probability measure on the trajectory space C=C([0,1];D‾)C = C([0,1];\overline{\mathcal{D}}), which we call a path measure. For noise scales 0<σ0≪σ10 < \sigma_0 \ll\sigma_1, define the variance-exploding schedule

σ(t):=σ01−tσ1t,g(t):=(ddtσ(t)2)1/2=σ(t)(2log⁡σ1σ0)1/2.(5)\sigma(t) := \sigma_0^{1-t}\sigma_1^t,\qquad g(t) := \left(\frac{\mathrm{d}}{\mathrm{d}t}\sigma(t)^2\right)^{1/2} = \sigma(t)\left(2\log\frac{\sigma_1}{\sigma_0}\right)^{1/2}. \tag*{(5)}

For any initial distribution μ0\mu_0, the corresponding RVE SDE is

dXt=g(t) dWt+dLt,X0∼μ0,0≤t≤1.(6)\mathrm{d}X_t = g(t)\,\mathrm{d}W_t + \mathrm{d}L_t,\qquad X_0 \sim\mu_0,\qquad0 \le t \le1. \tag*{(6)}

As the reference process, we use the RVE SDE (6) initialized from the uniform distribution μ0=ν\mu_0 = \nu, and denote the induced path measure by P→\overrightarrow{\mathbb{P}}. The following proposition justifies the use of the RVE SDE (6) as a noising process by showing that its marginals remain explicitly known when initialized from the uniform distribution, while, for an arbitrary initial distribution, the terminal law is approximately uniform.

Proposition 3.1.1. For the reflected noising process in (6), under the assumptions of Proposition C.3.1, the uniform distribution is invariant:

X0∼ν⟹Xt∼ν,0≤t≤1.X_0 \sim\nu\quad\Longrightarrow\quad X_t \sim\nu,\qquad0 \le t \le1.

Moreover, fix any initial probability distribution μ0\mu_0 on D‾\overline{\mathcal{D}}. For each n∈Nn \in\mathbb{N}, let X(n)X^{(n)} solve (6) with initial law μ0\mu_0 and noise schedule gn(⋅)g_n(\cdot) corresponding to scales 0<σ0,n<σ1,n0 < \sigma_{0,n} < \sigma_{1,n}, and denote its terminal law by μ1(n):=Law⁡(X1(n))\mu_1^{(n)} := \operatorname{Law}(X_1^{(n)}). Suppose that ∫01gn2(t) dt=σ1,n2−σ0,n2→n→∞∞\int_0^1 g_n^2(t)\,\mathrm{d}t = \sigma_{1,n}^2 - \sigma_{0,n}^2 \xrightarrow{n\to\infty} \infty. Then

∥μ1(n)−ν∥TV→n→∞0.\lVert\mu_1^{(n)} - \nu\rVert_{\mathrm{TV}} \xrightarrow{n\to\infty} 0.

Full formulation of Proposition 3.1.1 with proofs is available in Proposition C.3.1. In practice one should choose σ0≪σ1\sigma_0 \ll\sigma_1 to control the discrepancy between the terminal distribution μ1\mu_1 of X1X_1 and the uniform distribution ν\nu.

Consequently, since the initial marginal of P→\overrightarrow{\mathbb{P}} is uniform, (1) can be written as an expectation under the reference path measure Z=EP→[r(X0)/ν(X0)]Z = \mathbb{E}_{\overrightarrow{\mathbb{P}}}[r(X_0)/\nu(X_0)]. When used directly, it corresponds to uniform Monte Carlo. This estimator can have large variance when the mass of r(x)r(x) is concentrated in a small region of D\mathcal{D}, since most samples from the uniform distribution then make only a negligible contribution to the estimate. This motivates replacing the reference path measure with a more suitable proposal. Generally, any path measure Q→\overrightarrow{\mathbb{Q}}, which is equivalent to P→\overrightarrow{\mathbb{P}} (P→≪Q→\overrightarrow{\mathbb{P}} \ll\overrightarrow{\mathbb{Q}} and Q→≪P→\overrightarrow{\mathbb{Q}} \ll\overrightarrow{\mathbb{P}}), gives the importance weight

References

  1. [1]Akhound-Sadegh, T., Rector-Brooks, J., Bose, J., Mittal, S., Lemos, P., Liu, C.-H., Sendera, M., Ravanbakhsh, S., Gidel, G., Bengio, Y., Malkin, N., and Tong, A. Iterated denoising energy matching for sampling from boltzmann densities. In Forty-first International Conference on Machine Learning, 2024. URL https://openreview.net/forum?id=gVJmWlDF0Q.
  2. [2]Alfaro, M., Hilhorst, D., and Matano, H. The singular limit of the allen–cahn equation and the fitzhugh–nagumo system. Journal of Differential Equations, 245(2):505–565, 2008.arxiv.org/abs/0709.4071
  3. [4]Berner, J., Richter, L., Sendera, M., Rector-Brooks, J., and Malkin, N. From discrete-time policies to continuous-time diffusion samplers: Asymptotic equivalences and faster training. Transactions on Machine Learning Research, 2026. ISSN 2835-8856. URL https://openreview.net/forum?id=xLE3xJUudO. Expert Certification.arxiv.org/abs/2501.06148
  4. [5]Blessing, D., Jia, X., Esslinger, J., Vargas, F., and Neumann, G. Beyond ELBOs: A large-scale evaluation of variational methods for sampling. International Conference on Machine Learning (ICML), 2024.arxiv.org/abs/2406.07423
  5. [6]Bonciani, R., Del Duca, V., Frellesvig, H., Henn, J. M., Moriello, F., and Smirnov, V. A. Two-loop planar master integrals for higgs → 3 partons with full heavy-quark mass dependence. Journal of High Energy Physics, 2016(12), December 2016. ISSN 1029-8479. doi: 10.1007/jhep12(2016)096. URL http://dx.doi.org/10.1007/JHEP12(2016)096.
  6. [7]Borowka, S., Heinrich, G., Jahn, S., Jones, S., Kerner, M., Schlenk, J., and Zirke, T. pysecdec: A toolbox for the numerical evaluation of multi-scale integrals. Computer Physics Communications, 222:313–326, January 2018. ISSN 0010-4655. doi: 10.1016/j.cpc.2017.09.015. URL http://dx.doi.org/10.1016/j.cpc.2017.09.015.
  7. [8]Briol, F.-X., Oates, C. J., Girolami, M., Osborne, M. A., and Sejdinovic, D. Probabilistic integration. Statistical Science, 34(1):1–22, 2019.
  8. [9]Burdzy, K., Chen, Z.-Q., and Sylvester, J. The heat equation and reflected brownian motion in time-dependent domains. The Annals of Probability, 32(1B):775–804, 2004.
  9. [10]Cattiaux, P. Time reversal of diffusion processes with a boundary condition. Stochastic processes and their Applications, 28(2):275–292, 1988.DOI
  10. [11]Chen, Y., Georgiou, T. T., and Pavon, M. On the relation between optimal transport and schrödinger bridges: A stochastic control viewpoint. Journal of Optimization Theory and Applications, 169(2):671–691, 2016.arxiv.org/abs/1412.4430
  11. [12]Chen, Y., Georgiou, T. T., and Pavon, M. Stochastic control liaisons: Richard sinkhorn meets gaspard monge on a schrodinger bridge. Siam Review, 63(2):249–313, 2021.DOI
  12. [13]Deutschmann, N. and Götz, N. Accelerating hep simulations with neural importance sampling. Journal of High Energy Physics, 2024(3):83, 2024.arxiv.org/abs/2401.09069
  13. [14]Dick, J., Kuo, F. Y., and Sloan, I. H. High-dimensional integration: the quasi-monte carlo way. Acta Numerica, 22:133–288, 2013.DOI
  14. [15]Dinh, L., Krueger, D., and Bengio, Y. NICE: Non-linear independent components estimation. arXiv preprint arXiv:1410.8516, 2014.DOI
  15. [16]Fishman, N., Klarner, L., Bortoli, V. D., Mathieu, E., and Hutchinson, M. J. Diffusion models for constrained domains. Transactions on Machine Learning Research, 2023. ISSN 2835-8856. URL https://openreview.net/forum?id=xuWTFQ4VGO. Expert Certification.arxiv.org/abs/2304.05364
  16. [17]Foote, R. L. Regularity of the distance function. Proceedings of the American Mathematical Society, 92(1):153–155, 1984.DOI
  17. [18]Genz, A. Testing multidimensional integration routines. In Proc. of international conference on Tools, methods and languages for scientific and engineering computation, pp. 81–94, 1984.
  18. [19]Glasserman, P. Monte Carlo methods in financial engineering, volume 53. Springer, 2004.DOI
  19. [20]Gunter, T., Osborne, M. A., Garnett, R., Hennig, P., and Roberts, S. J. Sampling for inference in probabilistic models with fast bayesian quadrature. Advances in neural information processing systems, 27, 2014.arxiv.org/abs/1411.0439
  20. [21]Guo, W., Tao, M., and Chen, Y. Complexity analysis of normalizing constant estimation: from jarzynski equality to annealed importance sampling and beyond. In The Fourteenth International Conference on Learning Representations, 2026. URL https://openreview.net/forum?id=96fJALwotm.arxiv.org/abs/2502.04575
  21. [22]Hammersley, J. M. and Handscomb, D. C. Monte Carlo Methods. Methuen & Co. London; John Wiley & Sons, New York, 1964.DOI
  22. [23]Havens, A. J., Miller, B. K., Yan, B., Domingo-Enrich, C., Sriram, A., Levine, D. S., Wood, B. M., Hu, B., Amos, B., Karrer, B., Fu, X., Liu, G.-H., and Chen, R. T. Q. Adjoint sampling: Highly scalable diffusion samplers via adjoint matching. In Forty-second International Conference on Machine Learning, 2025. URL https://openreview.net/forum?id=6Eg10rHmg2.arxiv.org/abs/2504.11713
  23. [24]Holk, A., Strauch, C., and Trottner, L. Statistical guarantees for denoising reflected diffusion models. Journal of Machine Learning Research, 27(116):1–50, 2026.
  24. [25]Kaarnioja, V. Explicit solutions of genz test integrals. Applied Mathematics Letters, 163:109444, 2025.DOI
  25. [26]Kass, R. E. and Raftery, A. E. Bayes factors. Journal of the american statistical association, 90(430):773–795, 1995.
  26. [27]Kloek, T. and Van Dijk, H. K. Bayesian estimates of equation system parameters: an application of integration by monte carlo. Econometrica: Journal of the Econometric Society, pp. 1–19, 1978.DOI
  27. [28]Kong, Z. and Chaudhuri, K. The expressive power of a class of normalizing flow models. In International conference on artificial intelligence and statistics, pp. 3599–3609. PMLR, 2020.arxiv.org/abs/2006.00392
  28. [29]Kool, W., van Hoof, H., and Welling, M. Buy 4 REINFORCE samples, get a baseline for free! In ICLR 2019 Workshop: Deep Reinforcement Learning Meets Structured Prediction, 2019. URL https://openreview.net/forum?id=r1IgTGL5DE.
  29. [31]Le Gall, J.-F. Brownian motion, martingales, and stochastic calculus, volume 274. Springer, 2016.
  30. [32]LeCun, Y., Bottou, L., Bengio, Y., and Haffner, P. Gradient-based learning applied to document recognition. Proceedings of the IEEE, 86(11):2278–2324, 1998.
  31. [33]Liu, G.-H., Chen, T., Theodorou, E., and Tao, M. Mirror diffusion models for constrained and watermarked generation. Advances in Neural Information Processing Systems, 36:42898–42917, 2023.arxiv.org/abs/2310.01236
  32. [34]Loper, J. Uniform ergodicity for brownian motion in a bounded convex set. Journal of Theoretical Probability, 33(1):22–35, 2020.
  33. [35]Lou, A. and Ermon, S. Reflected diffusion models. In International Conference on Machine Learning, pp. 22675–22701. PMLR, 2023.arxiv.org/abs/2304.04740
  34. [36]L’Ecuyer, P. Quasi-monte carlo methods with applications in finance. Finance and Stochastics, 13 (3):307–349, 2009.DOI
  35. [37]Midgley, L. I., Stimper, V., Simm, G. N. C., Schölkopf, B., and Hernández-Lobato, J. M. Flow annealed importance sampling bootstrap. In The Eleventh International Conference on Learning Representations, 2023. URL https://openreview.net/forum?id=XCTVFJwS9LJ.arxiv.org/abs/2208.01893
  36. [38]Müller, T., McWilliams, B., Rousselle, F., Gross, M., and Novák, J. Neural importance sampling. ACM Transactions on Graphics (ToG), 38(5):1–19, 2019.
  37. [39]Nicoli, K. A., Anders, C. J., Funcke, L., Hartung, T., Jansen, K., Kessel, P., Nakajima, S., and Stornati, P. Estimation of thermodynamic observables in lattice field theories with deep generative models. Physical review letters, 126(3):032001, 2021.DOI
  38. [40]Niederreiter, H. Random number generation and quasi-Monte Carlo methods. SIAM, 1992.
  39. [42]O’Hagan, A. Bayes–hermite quadrature. Journal of statistical planning and inference, 29(3):245–260, 1991.DOI
  40. [43]Ott, K., Tiemann, M., Hennig, P., and Briol, F.-X. Bayesian numerical integration with neural networks. In Uncertainty in Artificial Intelligence, pp. 1606–1617. PMLR, 2023.arxiv.org/abs/2305.13248
  41. [44]Owen, A. B. Monte carlo variance of scrambled net quadrature. SIAM Journal on Numerical Analysis, 34(5):1884–1910, 1997.DOI
  42. [45]Pilipenko, A. An introduction to stochastic differential equations with reflection. Universitätsverlag Potsdam, 2014.
  43. [46]Rasmussen, C. E. and Ghahramani, Z. Bayesian monte carlo. Advances in neural information processing systems, pp. 505–512, 2003.
  44. [47]Rezende, D. and Mohamed, S. Variational inference with normalizing flows. In International conference on machine learning, pp. 1530–1538. PMLR, 2015.arxiv.org/abs/1505.05770
  45. [48]Schuss, Z. Brownian dynamics at boundaries and interfaces, volume 758. Springer, 2015.
  46. [49]Sendera, M., Kim, M., Mittal, S., Lemos, P., Scimeca, L., Rector-Brooks, J., Adam, A., Bengio, Y., and Malkin, N. Improved off-policy training of diffusion samplers. Advances in Neural Information Processing Systems, 37:81016–81045, 2024.
  47. [50]Song, Y., Sohl-Dickstein, J., Kingma, D. P., Kumar, A., Ermon, S., and Poole, B. Score-based generative modeling through stochastic differential equations. In International Conference on Learning Representations, 2021. URL https://openreview.net/forum?id=PxTIG12RRHS.
  48. [51]Taraldsen, G. Optimal learning from the doob-dynkin lemma. arXiv preprint arXiv:1801.00974, 2018.arxiv.org/abs/1801.00974
  49. [52]Vargas, F., Grathwohl, W. S., and Doucet, A. Denoising diffusion samplers. In International Conference on Learning Representations, 2023. URL https://openreview.net/forum?id=8ypnftTabu1f.arxiv.org/abs/2302.13834
  50. [53]Vargas, F., Padhy, S., Blessing, D., and Nüsken, N. Transport meets variational inference: Controlled monte carlo diffusions. In The Twelfth International Conference on Learning Representations, 2024. URL https://openreview.net/forum?id=PP1rudnxiW.arxiv.org/abs/2307.01050
  51. [54]Virtanen, P., Gommers, R., Oliphant, T. E., Haberland, M., Reddy, T., Cournapeau, D., Burovski, E., Peterson, P., Weckesser, W., Bright, J., et al. Scipy 1.0: fundamental algorithms for scientific computing in python. Nature methods, 17(3):261–272, 2020.
  52. [55]Warren, H. and Ramos, F. Fast fourier bayesian quadrature. In International Conference on Artificial Intelligence and Statistics, pp. 4555–4563. PMLR, 2024.
  53. [56]Williams, R. On time-reversal of reflected brownian motions. In Seminar on Stochastic Processes, 1987, pp. 265–276. Springer, 1987.DOI
  54. [57]Williams, R. J. Simple statistical gradient-following algorithms for connectionist reinforcement learning. Machine learning, 8(3):229–256, 1992.
  55. [58]Wirnsberger, P., Ballard, A. J., Papamakarios, G., Abercrombie, S., Racanière, S., Pritzel, A., Jimenez Rezende, D., and Blundell, C. Targeted free energy estimation via learned mappings. The Journal of Chemical Physics, 153(14), 2020.
  56. [59]Zhang, Q. and Chen, Y. Path integral sampler: A stochastic control approach for sampling. In International Conference on Learning Representations, 2022. URL https://openreview.net/forum?id=_uCb2ynRu7Y.

Paper details

vlnc:W_kgba6xqz
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 1

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Computer Science

Not recorded for this paper

View statement
Looking for related papers…
Contents