Pith. sign in

REVIEW 7 cited by

Sharp convergence rates for Langevin dynamics in the nonconvex setting

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 1805.01648 v4 pith:HYV3KU7F submitted 2018-05-04 stat.ML cs.LGmath.PRstat.CO

classification stat.MLcs.LGmath.PRstat.CO
keywords epsilonlangevincomplexitydistributioniterationleftmcmcright
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
abstract

We study the problem of sampling from a distribution $p^*(x) \propto \exp\left(-U(x)\right)$, where the function $U$ is $L$-smooth everywhere and $m$-strongly convex outside a ball of radius $R$, but potentially nonconvex inside this ball. We study both overdamped and underdamped Langevin MCMC and establish upper bounds on the number of steps required to obtain a sample from a distribution that is within $\epsilon$ of $p^*$ in $1$-Wasserstein distance. For the first-order method (overdamped Langevin MCMC), the iteration complexity is $\tilde{\mathcal{O}}\left(e^{cLR^2}d/\epsilon^2\right)$, where $d$ is the dimension of the underlying space. For the second-order method (underdamped Langevin MCMC), the iteration complexity is $\tilde{\mathcal{O}}\left(e^{cLR^2}\sqrt{d}/\epsilon\right)$ for an explicit positive constant $c$. Surprisingly, the iteration complexity for both these algorithms is only polynomial in the dimension $d$ and the target accuracy $\epsilon$. It is exponential, however, in the problem parameter $LR^2$, which is a measure of non-log-concavity of the target distribution.

Discussion (0). Sign in to comment.

Forward citations

Cited by 7 Pith papers

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

  1. Provable Quantum Speedups for Reaction-Rate Estimation in High-Dimensional Fokker-Planck Dynamics

    quant-ph 2026-01 conditional novelty 8.0 of 10

    A quantum algorithm estimates Fokker-Planck reaction rates with sublinear-time, polynomial-in-particle-number cost, giving an exponential-in-particle-number separation from the sharpest classical worst-case Langevin bounds.

  2. Generalization of Gibbs and Langevin Monte Carlo Algorithms in the Interpolation Regime

    cs.LG 2025-10 conditional novelty 7.0 of 10

    New PAC-Bayes bounds for the Gibbs posterior remain non-vacuous in the interpolation regime and can be approximated by Langevin Monte Carlo, but the tight experimental numbers rely on an unproved random-label calibrat...

  3. Multimodal sampling via Schr\"odinger-F\"ollmer samplers with temperatures

    math.NA 2025-12 conditional novelty 6.0 of 10

    Euler-discretized Schrödinger–Föllmer samplers with a temperature parameter provably converge at order O(h) in L2-Wasserstein distance, and high temperatures markedly improve multimodal sampling in experiments.

  4. An explicit splitting SAV scheme for the kinetic Langevin dynamics

    math.NA 2025-09 conditional novelty 6.0 of 10

    An explicit SSAV discretization of kinetic Langevin dynamics achieves order-one strong and weak convergence with error constants polynomial in time, even for superlinear-gradient potentials.

  5. Regime-Switching Langevin Monte Carlo Algorithms

    stat.CO 2025-08 conditional novelty 6.0 of 10

    Regime-switching LMC and KLMC variants inherit Gibbs invariance and get W2 convergence bounds; the headline FRS-KLMC O(1/sqrt(epsilon)) complexity is not supported by the given proof.

  6. kTULA: A Langevin sampling algorithm with improved KL bounds under super-linear log-gradients

    math.ST 2025-06 accept novelty 6.0 of 10

    kTULA achieves a non-asymptotic KL convergence of order lambda^(2-epsilon) for non-log-concave targets with super-linear log-gradients under a Log-Sobolev inequality, improving prior order-lambda KL bounds.

  7. Tamed Stochastic Gradient Hamiltonian Monte Carlo

    math.OC 2026-07 conditional novelty 5.0 of 10

    tSGHMC provably samples from strongly convex targets with superlinear, discontinuous stochastic gradients at a λ^{1/4} Wasserstein-2 rate.

Pith tools