mlsanityData to 7 Oct 2026

Score-Based Diffusion meets Annealed Importance Sampling

arXiv 2208.07698 ↗First posted 16 Aug 2022stat.ML cs.LGcited by 51

Not on the map. Why?

Cited by 22 in Diffusion models, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

More than twenty years after its introduction, Annealed Importance Sampling (AIS) remains one of the most effective methods for marginal likelihood estimation. It relies on a sequence of distributions interpolating between a tractable initial distribution and the target distribution of interest which we simulate from approximately using a non-homogeneous Markov chain. To obtain an importance sampling estimate of the marginal likelihood, AIS introduces an extended target distribution to reweight the Markov chain proposal. While much effort has been devoted to improving the proposal distribution used by AIS, an underappreciated issue is that AIS uses a convenient but suboptimal extended target distribution. We here leverage recent progress in score-based generative modeling (SGM) to approximate the optimal extended target distribution minimizing the variance of the marginal likelihood estimate for AIS proposals corresponding to the discretization of Langevin and Hamiltonian dynamics. We demonstrate these novel, differentiable, AIS procedures on a number of synthetic benchmark distributions and variational auto-encoders.

Key facts. Cited by 51 papers in the map (data through 7 Oct 2026). Builds on 22 of its 56 references that are in the map. First posted on arXiv 16 Aug 2022.

Builds on

22 of 56 references.

References · 22

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  3. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 8,964
  4. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,005
  5. 2016Layer Normalizationcited by 5,579
  6. 2014NICE: Non-linear Independent Components Estimationcited by 1,052
  7. 2021Maximum Likelihood Training of Score-Based Diffusion Modelscited by 670
  8. 2021Diffusion Schr"odinger Bridge with Applications to Score-Based Generative Modelingcited by 594
  9. 2015Importance Weighted Autoencoderscited by 454
  10. 2021Score-Based Generative Modeling with Critically-Damped Langevin Diffusioncited by 230
  11. 2020Stochastic Normalizing Flowscited by 156
  12. 2021Annealed Flow Transport Monte Carlocited by 90
10 more
  1. 2022Continual Repeated Annealed Flow Transport Monte Carlocited by 66
  2. 2021MCMC Variational Inference via Uncorrected Hamiltonian Annealingcited by 39
  3. 2021Monte Carlo Variational Auto-Encoderscited by 36
  4. 2021Differentiable Annealed Importance Sampling and the Perils of Gradient Noisecited by 36
  5. 2020An invitation to sequential Monte Carlo samplerscited by 33
  6. 2019The Thermodynamic Variational Objectivecited by 26
  7. 2018Improving Explorability in Variational Inference with Annealed Variational Objectivescited by 24
  8. 2021Surrogate Likelihoods for Variational Annealed Importance Samplingcited by 14
  9. 2021q-Paths: Generalizing the Geometric Annealing Path using Power Meanscited by 12
  10. 2021Simulated annealing from continuum to discretization: a convergence analysis via the Eyring--Kramers lawcited by 5

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 2; 2023: 12; 2024: 15; 2025: 16; 2026: 6.

51 papers.

Sort:

By area:

  1. 2022An optimal control perspective on diffusion-based generative modelingcited by 145
  2. 2023Denoising Diffusion Samplerscited by 144
  3. 2024Iterated Denoising Energy Matching for Sampling from Boltzmann Densitiescited by 113
  4. 2023Improved sampling via learned diffusionscited by 103
  5. 2024DiffMOT: A Real-time Diffusion-based Multiple Object Tracker with Non-linear Predictioncited by 59
  6. 2023Diffusion Generative Flow Samplers: Improving learning signals through partial trajectory optimizationcited by 57
  7. 2024Beyond ELBOs: A Large-Scale Evaluation of Variational Methods for Samplingcited by 55
  8. 2025DIME:Diffusion-Based Maximum Entropy Reinforcement Learningcited by 52
  9. 2024Sequential Controlled Langevin Diffusionscited by 47
  10. 2025Underdamped Diffusion Bridges with Applications to Samplingcited by 28
  11. 2024Stochastic Localization via Iterative Posterior Samplingcited by 27
  12. 2024Liouville Flow Importance Samplercited by 25
38 more
  1. 2025No Trick, No Treat: Pursuits and Challenges Towards Simulation-free Training of Neural Samplerscited by 25
  2. 2025Trust Region Constrained Measure Transport in Path Space for Stochastic Optimal Control and Inferencecited by 22
  3. 2023Transport meets Variational Inference: Controlled Monte Carlo Diffusionscited by 21
  4. 2025MDNS: Masked Diffusion Neural Sampler via Stochastic Optimal Controlcited by 20
  5. 2025From discrete-time policies to continuous-time diffusion samplers: Asymptotic equivalences and faster trainingcited by 17
  6. 2025Outsourced diffusion sampling: Efficient posterior inference in latent spaces of generative modelscited by 16
  7. 2025RNE: plug-and-play diffusion inference-time control and energy-based trainingcited by 16
  8. 2023Efficient Training of Energy-Based Models Using Jarzynski Equalitycited by 15
  9. 2024BNEM: A Boltzmann Sampler Based on Bootstrapped Noised Energy Matchingcited by 14
  10. 2023Entropy-based Training Methods for Scalable Neural Implicit Samplercited by 13
  11. 2025Complexity Analysis of Normalizing Constant Estimation: from Jarzynski Equality to Annealed Importance Sampling and beyondcited by 11
  12. 2023Transport with Support: Data-Conditional Diffusion Bridgescited by 9
  13. 2024Markovian Flow Matching: Accelerating MCMC with Continuous Normalizing Flowscited by 9
  14. 2024Learned Reference-based Diffusion Sampling for multi-modal distributionscited by 9
  15. 2023Adaptive Annealed Importance Sampling with Constant Rate Progresscited by 7
  16. 2024What's the score? Automated Denoising Score Matching for Nonlinear Diffusionscited by 7
  17. 2026Bridge Matching Sampler: Scalable Sampling via Generalized Fixed-Point Diffusion Matchingcited by 7
  18. 2023Differentiating Metropolis-Hastings to Optimize Intractable Densitiescited by 5
  19. 2025Reinforced sequential Monte Carlo for amortised samplingcited by 5
  20. 2024Ensemble-Based Annealed Importance Samplingcited by 4
  21. 2025End-To-End Learning of Gaussian Mixture Priors for Diffusion Samplercited by 4
  22. 2022Proposal of a Score Based Approach to Sampling Using Monte Carlo Estimation of Score and Oracle Access to Target Densitycited by 3
  23. 2024Critically Damped Third-Order Langevin Dynamicscited by 3
  24. 2026Sample-efficient evidence estimation of score based priors for model selectioncited by 3
  25. 2023Resampling Gradients Vanish in Differentiable Sequential Monte Carlo Samplerscited by 2
  26. 2023Reparameterized Variational Rejection Samplingcited by 2
  27. 2024Denoising Diffusion Variational Inference: Diffusion Models as Expressive Variational Posteriorscited by 2
  28. 2024Differentiable Annealed Importance Sampling Minimizes The Symmetrized Kullback-Leibler Divergence Between Initial and Target Distributioncited by 2
  29. 2026Learning Energy-Based Models from Stochastic Interpolants using Spatiotemporal Differencescited by 2
  30. 2023To smooth a cloud or to pin it down: Guarantees and Insights on Score Matching in Denoising Diffusion Modelscited by 1
  31. 2025Continuously Tempered Diffusion Samplerscited by 1
  32. 2025One-Step Diffusion Samplers via Self-Distillation and Deterministic Flowcited by 1
  33. 2026Composing diffusion priors with explicit physical context via generative Gibbs samplingcited by 1
  34. 2024Control, Transport and Sampling: Towards Better Loss Designcited by 0
  35. 2025Efficient Weighted Sampling via Score-based Generative Modelscited by 0
  36. 2025Jarzynski Reweighting and Sampling Dynamics for Training Energy-Based Models: Theoretical Analysis of Different Transition Kernelscited by 0
  37. 2025Analysis of kinetic Langevin Monte Carlo under the stochastic exponential Euler discretization from underdamped all the way to overdampedcited by 0
  38. 2026Total Variation Guarantees for Sampling with Stochastic Localizationcited by 0

50 most cited of 51 All 51 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2022Sampling is as easy as learning the score: theory for diffusion models with minimal data assumptions
  2. 2025From Score Matching to Diffusion: A Fine-Grained Error Analysis in the Gaussian Setting
  3. 2025Efficient and Unbiased Sampling from Boltzmann Distributions via Variance-Tuned Diffusion Models
  4. 2025Scalable Discrete Diffusion Samplers: Combinatorial Optimization and Statistical Physics
  5. 2024A Practical Diffusion Path for Sampling
  6. 2025Non-asymptotic Analysis of Diffusion Annealed Langevin Monte Carlo for Generative Modelling
  7. 2025Efficient Approximate Posterior Sampling with Annealed Langevin Monte Carlo
  8. 2023A Complete Recipe for Diffusion Generative Models