mlsanityData to 14 Oct 2026

Debiasing Guidance for Discrete Diffusion with Sequential Monte Carlo

arXiv 2502.06079 ↗First posted 10 Feb 2025cs.LGcited by 31

On the map: Diffusion models → Diffusion foundations and sampling

Abstract (arXiv)

Discrete diffusion models are a class of generative models that produce samples from an approximated data distribution within a discrete state space. Often, there is a need to target specific regions of the data distribution. Current guidance methods aim to sample from a distribution with mass proportional to p0(x0) p(zeta|x0)α but fail to achieve this in practice. We introduce a Sequential Monte Carlo algorithm that generates unbiasedly from this target distribution, utilising the learnt unconditional and guided process. We validate our approach on low-dimensional distributions, controlled images and text generations. For text generation, our method provides strong control while maintaining low perplexity compared to guidance-based approaches.

Key facts. Cited by 31 papers in the map (data through 14 Oct 2026). Placed in Diffusion foundations and sampling, one of 14 directions in Diffusion models. Builds on 32 of its 43 references that are in the map. First posted on arXiv 10 Feb 2025.

Builds on

32 of 43 references.

Closest ancestors · 12

Same direction, newest first

  1. 2025Inference-Time Alignment in Diffusion Models with Reward-Guided Generation: Tutorial and Review16 Jan
  2. 2024What does guidance do? A fine-grained analysis in a simple setting19 Sep
  3. 2024Classifier-Free Guidance is a Predictor-Corrector16 Aug
  4. 2024Derivative-Free Guidance in Continuous and Discrete Diffusion Models with Soft Value-Based Decoding15 Aug
  5. 2024Amortizing intractable inference in diffusion models for vision, language, and control31 May
  6. 2023SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis4 Jul
  7. 2023Denoising Diffusion Samplers27 Feb
  8. 2022A Continuous Time Framework for Discrete Denoising Models30 May
  9. 2021High-Resolution Image Synthesis with Latent Diffusion Models20 Dec
  10. 2021Diffusion Models Beat GANs on Image Synthesis11 May
  11. 2020Score-Based Generative Modeling through Stochastic Differential Equations26 Nov
  12. 2020Denoising Diffusion Probabilistic Models19 Jun

Other references · 20

  1. 2019Generative Modeling by Estimating Gradients of the Data Distributioncited by 4,101
  2. 2023Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasetscited by 2,581
  3. 2020DeBERTa: Decoding-enhanced BERT with Disentangled Attentioncited by 2,181
  4. 2016Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translationcited by 1,475
  5. 2022Building Normalizing Flows with Stochastic Interpolantscited by 995
  6. 2023Stochastic Interpolants: A Unifying Framework for Flows and Diffusionscited by 939
  7. 2022Equivariant Diffusion for Molecule Generation in 3Dcited by 677
  8. 2024Simplified and Generalized Masked Diffusion for Discrete Datacited by 626
  9. 2022DiGress: Discrete Denoising diffusion for graph generationcited by 520
  10. 2023Directly Fine-Tuning Diffusion Models on Differentiable Rewardscited by 448
  11. 2021DExperts: Decoding-Time Controlled Text Generation with Experts and Anti-Expertscited by 441
  12. 2024Generative Flows on Discrete State-Spaces: Enabling Multimodal Flows with Applications to Protein Co-Designcited by 305
8 more
  1. 2024Adjoint Matching: Fine-tuning Flow and Diffusion Generative Models with Memoryless Stochastic Optimal Controlcited by 233
  2. 2023Practical and Asymptotically Exact Conditional Sampling in Diffusion Modelscited by 232
  3. 2023Likelihood-Based Diffusion Language Modelscited by 151
  4. 2024Unlocking Guidance for Discrete State-Space Diffusion and Flow Modelscited by 129
  5. 2024NETS: A Non-Equilibrium Transport Samplercited by 69
  6. 2023Transport meets Variational Inference: Controlled Monte Carlo Diffusionscited by 21
  7. 2023Efficient Training of Energy-Based Models Using Jarzynski Equalitycited by 15
  8. 2024Training-Free Guidance for Discrete Diffusion Models for Molecular Generationcited by 6

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 18; 2026: 13.

31 papers.

Sort:

By area:

  1. 2025Feynman-Kac Correctors in Diffusion: Annealing, Guidance, and Product of Expertscited by 96
  2. 2025DriftLite: Lightweight Drift Control for Inference-Time Scaling of Diffusion Modelscited by 23
  3. 2025Composition and Control with Distilled Energy Diffusion Models and Sequential Monte Carlocited by 20
  4. 2025Split Gibbs Discrete Diffusion Posterior Samplingcited by 18
  5. 2025Solving Inverse Problems via Diffusion-Based Priors: An Approximation-Free Ensemble Sampling Approachcited by 18
  6. 2025RNE: plug-and-play diffusion inference-time control and energy-based trainingcited by 17
  7. 2025Inference-Time Scaling of Discrete Diffusion Models via Importance Weighting and Optimal Proposal Designcited by 14
  8. 2025Discrete Neural Flow Samplers with Locally Equivariant Transformercited by 13
  9. 2025CREPE: Controlling Diffusion with Replica Exchangecited by 11
  10. 2026Discrete Feynman-Kac Correctorscited by 11
  11. 2025Steering Generative Models with Experimental Data for Protein Fitness Optimizationcited by 7
  12. 2025Conditional Diffusion Models with Classifier-Free Gibbs-like Guidancecited by 5
19 more
  1. 2025Constrained Diffusion for Protein Design with Hard Structural Constraintscited by 5
  2. 2026Contrastive Distribution Matching for Amortized Sequential Monte Carlo in Discrete Diffusioncited by 5
  3. 2025ProSpero: Active Learning for Robust Protein Design Beyond Wild-Type Neighborhoodscited by 3
  4. 2025PriorGuide: Test-Time Prior Adaptation for Simulation-Based Inferencecited by 3
  5. 2025Foundations of Diffusion Models in General State Spaces: A Self-Contained Introductioncited by 2
  6. 2026Compression as Adaptation: Implicit Visual Representation with Diffusion Foundation Modelscited by 2
  7. 2026Simple Approximation and Derivative Free Inference-Time Scaling for Diffusion Models via Sequential Monte Carlo on Path Measurescited by 2
  8. 2025Improving Text Style Transfer using Masked Diffusion Language Models with Inference-time Scalingcited by 1
  9. 2025Deterministic Discrete Denoisingcited by 1
  10. 2026D5P4: Partition Determinantal Point Process for Diversity in Parallel Discrete Diffusion Decodingcited by 1
  11. 2026BFMT: Enhancing Search Capabilities of Tree Sampler via Bootstrap Flow-Map Treecited by 1
  12. 2026FluxLite: Inference-Time Proposal Control for Discrete Diffusion Modelscited by 1
  13. 2025Efficient Inference for Coupled Hidden Markov Models in Continuous Time and Discrete Spacecited by 0
  14. 2026Focus on the Core: Empowering Diffusion Large Language Models by Self-Contrastcited by 0
  15. 2026Sequentially-Controlled Interactive Multi-Particle Flow-Maps for Online Feedback-Driven Searchcited by 0
  16. 2026LATS: Levy Adaptive Tree Sampling for Feedback-Driven Diverse Target Discoverycited by 0
  17. 2026Parallel Tempering for Diffusion-Based Combinatorial Optimizationcited by 0
  18. 2026Risk-Calibrated Proposal Transport for Finite-Particle Diffusion Steeringcited by 0
  19. 2026Learning ω-Powered Conditional Distributions with Importance-Weighted Flow Matchingcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Multilevel and Sequential Monte Carlo for Training-Free Diffusion Guidance
  2. 2024Simple Guidance Mechanisms for Discrete Diffusion Models
  3. 2026On the Robustness of Distribution Support under Diffusion Guidance
  4. 2026Generating from Discrete Distributions Using Diffusions: Insights from Random Constraint Satisfaction Problems
  5. 2025Fast Solvers for Discrete Diffusion Models: Theory and Applications of High-Order Algorithms
  6. 2025Discrete Guidance Matching: Exact Guidance for Discrete Flow Matching
  7. 2025Constrained Discrete Diffusion
  8. 2026Discrete Diffusion Inference-Time Control with Nested Sequential Monte Carlo