Discrete Diffusion Sampling

Latest papers

20 papers1 paper in the last four weeks, with none the four weeks before. 0.0% of all new papers.

Online Class Posteriors for Interpretable and Controllable Discrete Diffusion
Nov 2, 2026Felix Koulischer, Gianni Van De Velde, François Remy +2Discrete Diffusion Sampling Diffusion Model Guidance
Ghent University - imec · Parallia Healthcare
An ODE-based Sampler for Discrete Diffusion Models
Oct 29, 2026Samson Gourevitch, Yazid Janati, Eric Moulines +3Discrete Diffusion Sampling Diffusion Model Sampling
CMAP, Ecole polytechnique · Institute of Foundation Models · MBZUAI +2
The information geometry of product-reference discrete diffusion: Interaction growth complexity and optimal scheduling
Aug 28, 2026Martin J. WainwrightDiscrete Diffusion Sampling Diffusion Sampling
Lab for Information and Decision Systems Statistics and Data Science Center EECS and Mathematics, Massachusetts Institute of Technology
LESS Is More: Mutual-Stability Sampling for Diffusion Language Models
Jun 15, 2026Amr Mohamed, Guokan Shang, Michalis VazirgiannisMasked Diffusion Models Diffusion Language Model Decoding
1MBZUAI · 2Ecole Polytechnique
teasr: training-efficient any-step diffusion transformer for real-world image super-resolution
Jun 15, 2026Xiang Gao, Chenxin Zhu, Yushun Fang +2Diffusion Transformer Diffusion Model Distillation
Shanghai Jiao Tong University, China
High-Fidelity Two-Step Image Generation via Teacher-Aligned End-to-End Distillation
Jun 10, 2026Dongyang Liu, Ruoyi Du, David Liu +7Diffusion Model Distillation Few-Step Diffusion Sampling
Z-Image Team, Alibaba Group
Efficient One-Step Diffusion Restoration Model with Compact Token Compression and Linear Attention
May 22, 2026Bingtian Qiao, Yue Shi, Yingjie Zhou +3Diffusion Models for Image Restoration Discrete Diffusion Sampling
Shanghai Jiao Tong University · Fuzhou University · Shanghai AI Laboratory
Forward-Learned Discrete Diffusion: Learning how to noise to denoise faster
May 18, 2026Grigory Bartosh, Teodora Pandeva, Sushrut Karmalkar +1Few-Step Diffusion Sampling Diffusion Model Acceleration
University of Amsterdam · Microsoft Research, Cambridge
Discrete Stochastic Localization for Non-autoregressive Generation
May 13, 2026Yunshu Wu, Jiayi Cheng, Longxuan Yu +4Non-Autoregressive Text Generation Masked Diffusion Models
University of California Riverside · New York University
Teacher-Feature Drifting: One-Step Diffusion Distillation with Pretrained Diffusion Representations
May 8, 2026Yuan Zhang, Chenyi Li, Guoqing Ma +7Diffusion Model Distillation Diffusion Sampling
JD Explore Academy, China · The Hong Kong University of Science and Technology · Tsinghua University +1
Consistent Diffusion Language Models
Apr 30, 2026Hasan Amin, Yuan Gao, Yaser Souri +4Few-Step Diffusion Sampling Diffusion Model Acceleration
Department of Computer Science, Purdue University · Microsoft
NI Sampling: Accelerating Discrete Diffusion Sampling by Token Order Optimization
Apr 20, 2026Enshu Liu, Xuefei Ning, Yu Wang +1Few-Step Diffusion Sampling Diffusion Language Model Decoding
Department of EE, Tsinghua University · Microsoft Research
Generalized Discrete Diffusion from Snapshots
Mar 22, 2026Oussama Zekri, Théo Uscidda, Nicolas Boullé +1Diffusion Models Discrete Diffusion Sampling
CREST, ENSAE, Institut Polytechnique de Paris, Paris, France · Department of Mathematics, Imperial College London, London, SW7 2AZ, UK
Beyond Single Tokens: Distilling Discrete Diffusion Models via Discrete MMD
Mar 20, 2026Emiel Hoogeboom, David Ruhe, Jonathan Heek +2Diffusion Model Distillation Diffusion Models
Google DeepMind Amsterdam
D5P4: Partition Determinantal Point Process for Diversity in Parallel Discrete Diffusion Decoding
Mar 19, 2026Jonathan Lys, Vincent Gripon, Axel Marmoret +4Diffusion Language Model Decoding Language Model Decoding
IMT Atlantique, Lab-STICC, UMR CNRS 6285, F-29238 Brest, France. · Sony Europe Ltd. Stuttgart Technology Center, EUREC, Germany.
DualTSR: Unified Dual-Diffusion Transformer for Scene Text Image Super-Resolution
Mar 15, 2026Axi Niu, Kang Zhang, Qingsen Yan +3Discrete Diffusion Sampling Image Super-Resolution
Northwestern Polytechnical University · Korea Advanced Institute of Science and Technology
The Diffusion Duality, Chapter II: $Ψ$ -Samplers
Feb 24, 2026Justin Deschenaux, Caglar Gulcehre, Subham Sekhar SahooDiffusion Models Diffusion Sampling
EPFL, Lausanne, Switzerland · Microsoft AI · Cornell Tech, NY
Discrete Stochastic Localization for Non-autoregressive Generation
Feb 18, 2026Yunshu Wu, Jiayi Cheng, Longxuan Yu +4Non-Autoregressive Text Generation Diffusion Language Models
University of California Riverside · New York University
Efficient Sampling with Discrete Diffusion Models: Sharp and Adaptive Guarantees
Feb 16, 2026Daniil Dmitriev, Zhihan Huang, Yuting WeiSample Complexity Diffusion Model Sampling
Department of Statistics and Data Science, the Wharton School, University of Pennsylvania
Improved Sampling Schedules for Discrete Diffusion Models
Feb 6, 2026Alberto Foresti, Mustapha Bounoua, Giulio Franzese +2Diffusion Sampling Discrete Diffusion Sampling
Department of Data Science, EURECOM, 450 Route des Chappes, 06410 Biot, France · Donders Institute for Brain, Cognition and Behaviour, Radboud University, Postbus 9102, 6500 HC Nijmegen, Netherlands

Nov 2, 2026J/K move · Enter open · S save

Online Class Posteriors for Interpretable and Controllable Discrete Diffusion

Felix Koulischer, Gianni Van De Velde, François Remy, Chris Develder+1

Ghent University - imec · Parallia Healthcare

Discrete Diffusion Sampling Diffusion Model Guidance Diffusion Guidance Conditional Diffusion Models Masked Diffusion Models

Open paper AskPDF arXiv

Discrete diffusion models remain far less understood than their continuous counterparts, despite recent breakthroughs across modalities, most notably in text. Leveraging tools from continuous diffusion, we formulate a theoretical framework to track the class-conditional posterior in an online manner by combining conditional and unconditional model predictions. In masked language modeling, this posterior offers highly interpretable dynamics: tokens aligned with a class boost the posterior, while contradictory ones suppress it. We also show that the posterior can be used for logical guidance operations (AND, OR, NOT) in discrete diffusion. These operations enable precise control over generation, such as avoiding specific classes. In addition, we discuss a mismatch in the formulation of classifier-free guidance between discrete and continuous settings: discrete models mix the data distributions, while continuous models mix the marginals. We validate our methods and discuss their potential applications on synthetic data, with all quantities analytically tractable, as well as on MNIST, text, and antibody sequences.