mlsanityData to 7 Oct 2026

Diffusion foundations and sampling

2,909 papers14 this weekMedium confidence in this grouping: mostly one topic, with blurry edgesMedium confidence in this grouping: mostly one topic, with blurry edges · About

Keywords: diffusion, training, generative, generation, sampling

Key facts. Diffusion foundations and sampling: 2,909 papers on arXiv since 2020 (data through 7 Oct 2026). 2,359 of its 2,909 mapped papers (81%) cite “Denoising Diffusion Probabilistic Models” (2020, arXiv 2006.11239). 1,781 of its 2,909 mapped papers (61%) cite “Score-Based Generative Modeling through Stochastic Differential Equations” (2020, arXiv 2011.13456). 83 new papers 17 Sep–7 Oct.

What’s new

17 Sep – 7 Oct 2026

83 new papers, 17 Sep – 7 Oct; 14 this week.

  1. This week · 14
  2. Enabling Preference-driven Unlearning in Few-step Distilled Text-to-Image Diffusion Models
  3. How Many Directions Must a Truncated Diffusion Sampler Retain? Matching Bounds Under Power-Law Spectra also fits: Flow matching, time series and science
  4. Koopman Observers for Diffusion Acceleration: Correcting Feature Forecasts with Shallow Measurements
  5. Universal Local Error and Realized Amplification for the First-Order EDM Predictor also fits: Flow matching, time series and science
  6. Think Before You Paint: Recursive Latent Reasoning for Diffusion Models also fits: Image generation, text-to-image and editing
  7. Consistent Distribution Matching for Data-Free Diffusion Distillation also fits: Flow matching, time series and science
  8. Lower Bounds for Parallel Diffusion Sampling
  9. Learning Transition Kernels of Jump-Diffusion Processes with Conditional Diffusion Models also fits: Flow matching, time series and science
  10. Feature Information Dynamics in Diffusion
  11. Two Halves are More than One: Phase-wise Velocity Distillation for Fast and High-Quality Image Generation also fits: Flow matching, time series and science
All 83
  1. Uniform Discrete Diffusion Models are Minimax Optimal for Estimating Distributions with Small Effective Support Size
  2. CentriQ: Calibration-Free Quantization of Diffusion Transformers via Exact Mean Centering
  3. CACFG: Curvature-Aware Classifier-Free Guidance and Optimal Control also fits: Diffusion and flow policies for robotics
  4. Usefulness of Quantile-Aware Diffusion Modeling for Highly Imbalanced Tabular Data also fits: Flow matching, time series and science
  5. Earlier · 69
  6. Should We Skip Diffusion? also fits: Inverse problems and image restoration
  7. Cross-Time Directional Selection in Diffusion Sampling also fits: Flow matching, time series and science
  8. Best-of-N Guidance for Test-time Diffusion Alignment
  9. D-DOIT: Training-free Adaptation of Discrete Diffusion via Doob's h-Transform
  10. Gradient-Free Sampling from Generative Models via Stochastic Bounded Extremum Seeking
  11. Discrete Diffusion for Large Graph Generation via Structural Candidate Restriction also fits: Flow matching, time series and science
  12. Probability flow ODEs in score-based and reflected diffusion models also fits: Flow matching, time series and science
  13. IDRF: Inverse-Distilled Reward Fine-tuning of Masked Discrete Diffusion Models
  14. Rethinking What to Cache in Few-Step Diffusion Transformers: Solver-Aware Target Selection
  15. Iterating Consistency Models: Stability, Error Bounds and Noise Schedules also fits: Flow matching, time series and science
  16. Adaptive Second-Order Solvers for Fast Stochastic Diffusion Sampling
  17. Controlling Polar Exposure to Delay Memorization in Diffusion Models
  18. Jumping up and down: Denoiser diffusion models for discrete ordinal data
  19. Correcting Guided Diffusion Trajectories with Spectral Alignment also fits: Inverse problems and image restoration
  20. Embedding Prediction Helps Image Generation also fits: Image generation, text-to-image and editing
  21. Sample complexity bounds for categorical Markov random fields via Discrete Diffusions also fits: Flow matching, time series and science
  22. Error-Corrected Inference-Time Scaling for Imperfect Diffusion Models
  23. Learned End-to-End Guidance Schedules for Diffusion Models
  24. Feature Selective Model Collapse in Diffusion Models: Total Replacement versus Fixed-Budget Training
  25. Joint Branch-Space Transform Coding for Diffusion Activation Quantization with Classifier-Free Guidance also fits: Flow matching, time series and science
  26. Triangular Resampling for Long-Horizon Motion Generation also fits: Flow matching, time series and science
  27. Rank-Aware Speculative Sampling for Diffusion Draft Trees also fits: Flow matching, time series and science
  28. Specificity-Aware Diffusion Steering via Variance-Reduced Sequential Monte Carlo also fits: Flow matching, time series and science
  29. Looped Diffusion Transformer
  30. From Modes to Memories: Characterizing the Scale-Space Dynamics of Diffusion Models
  31. MIND: Marginal-Invariant Neural Dependency Diffusion for Mixed-Type Tabular Generation also fits: Medical imaging and segmentation
  32. Why Do Conventional World Models Fail to Learn Cellular Automata? also fits: Flow matching, time series and science
  33. WEIRDO: WEak resIdual Regularized DOob's h-transform diffusion alignment
  34. CAST: Causal Advantage-Structured Training with Spatially Grounded Compositional Rewards for Diffusion Models also fits: Diffusion and flow policies for robotics
  35. CDMD: A Cross-Dataset Mixed-Type Diffusion Model for Tabular Data
  36. Visualizing Distribution Coverage in Generative Diffusion Models
  37. Distilling Diffusion Score Discrepancy for Efficient Training Data Attribution
  38. Adversarial Training for Pixel Diffusion also fits: Inverse problems and image restoration
  39. Improved Distributional Diffusion Models
  40. LDM-is-AE: Latent Diffusion Model is an Auto-Encoder for End-to-End Image Generation also fits: Inverse problems and image restoration
  41. Sharp Convergence and Sampling Trade-offs for Riemannian Diffusion under Nonnegative Ricci Curvature also fits: Flow matching, time series and science
  42. RA-CFGCache: From Branch-Level Criteria to Guided-Risk Control under Classifier-Free Guidance also fits: Flow matching, time series and science
  43. Representation by Design in Generation: Cross-View Class-Token Alignment in Diffusion Transformers also fits: Image generation, text-to-image and editing
  44. On the spectral properties of generative denoiser Jacobians
  45. FluxLite: Inference-Time Proposal Control for Discrete Diffusion Models also fits: Discrete and text diffusion
  46. PDMD: Projected Distribution Matching Distillation for Video Diffusion Models
  47. Simplex Diffusion Models
  48. First Learn, Then Memorize: The Spectral Bias of Diffusion Models
  49. Weighting Schedules Govern What and When Score-Based Generative Models Learn from Multimodal Data
  50. Scaffold Then Internalize: Representation Injection for Diffusion Transformers
  51. Learning to Re-Draft: A Variational Stackelberg Game for Discrete Diffusion also fits: Discrete and text diffusion
  52. Retrieval-Augmented Diffusion Modeling for Stochastic Discount Factor Portfolios
  53. SAGE: Subspace Alignment for Classifier-Free Guidance in Mixture-of-Experts Diffusion Models also fits: Flow matching, time series and science
  54. Residual-Stream Burden Shapes Representation Learning in Diffusion Transformers
  55. Chameleon: Dynamic Format Adapter for Efficient Diffusion
  56. Domain-Adapted Diffusion Models for Conditional Independence Testing
  57. A statistical theory for blind denoising: minimax estimation of the noise level from a single sample also fits: Inverse problems and image restoration
  58. DiffPTS: Rethinking Diffusion ELBO for Probabilistic Time Series Forecasting also fits: Flow matching, time series and science
  59. Continual Data Unlearning in Diffusion Models via Transition-based Regularization also fits: Discrete and text diffusion
  60. Active Data Acquisition with Side Information via Discrete Diffusion Priors also fits: Inverse problems and image restoration
  61. First-Order Stationarity of Reverse Diffusions also fits: Flow matching, time series and science
  62. Accelerating Video Diffusion via Training-Free Trajectory Routing also fits: Video generation
  63. Accurate Sampling from Diffusion Models also fits: Flow matching, time series and science
  64. From Graphs to Feeders: Constraint-Guided Diffusion for Rule-Compliant Feeder Generation also fits: Flow matching, time series and science
  65. On the Diffusibility of High-Dimensional Latents also fits: Flow matching, time series and science
  66. Robustness of Diffusion Models under Distribution Shift
  67. Discrete Diffusion Models via Evolving Variational Autoregressive Networks also fits: Flow matching, time series and science
  68. Double Descent and Malign Overfitting in Diffusion Models
  69. Rethinking Diffusion Segmentation: When Does It Rely on Its Noisy State, and Does Diffusion Matter?
  70. Leveraging Inference-Time Compute for Diffusion Models via Global Scheduling of Denoising Trajectories
  71. D-IMPL: A Diffusion-based Solver for Parameterized BBOs
  72. Schedule optimization for tau-leaping in masked discrete diffusion also fits: Flow matching, time series and science
  73. Quantization-Aware Kalman Estimation for Diffusion Sampling
  74. Limits of Confidence in Diffusion

Foundations

Most cited by papers here

  1. Denoising Diffusion Probabilistic Models2020cited by 2,359 here
  2. Score-Based Generative Modeling through Stochastic Differential Equations2020cited by 1,781 here
  3. High-Resolution Image Synthesis with Latent Diffusion Models2021cited by 1,697 here
  4. Denoising Diffusion Implicit Models2020cited by 1,438 here
  5. Diffusion Models Beat GANs on Image Synthesis2021cited by 1,409 here

Lineage

The papers this direction builds on, oldest first.

Each paper on the line cites the one before.

  1. 2020Denoising Diffusion Probabilistic Modelscited by 2,359 here
  2. 2020Denoising Diffusion Implicit Modelscited by 1,438 here
  3. 2021Improved Denoising Diffusion Probabilistic Modelscited by 772 here
  4. 2021Diffusion Models Beat GANs on Image Synthesiscited by 1,409 here
  5. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 1,697 here
  6. 2022Hierarchical Text-Conditional Image Generation with CLIP Latentscited by 607 here
  7. 2022Photorealistic Text-to-Image Diffusion Models with Deep Language Understandingcited by 719 here
  8. 2023Adding Conditional Control to Text-to-Image Diffusion Modelscited by 302 here
  9. 2023SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesiscited by 469 here
  10. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesiscited by 480 here
14 more it builds on, not in this chain
  1. 2020DiffWave: A Versatile Diffusion Model for Audio Synthesiscited by 363 here
  2. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 1,781 here
  3. 2021Variational Diffusion Modelscited by 355 here
  4. 2021GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Modelscited by 360 here
  5. 2022Video Diffusion Modelscited by 347 here
  6. 2022Elucidating the Design Space of Diffusion-Based Generative Modelscited by 884 here
  7. 2022DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Stepscited by 679 here
  8. 2022Classifier-Free Diffusion Guidancecited by 812 here
  9. 2022Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flowcited by 484 here
  10. 2022Flow Matching for Generative Modelingcited by 526 here
  11. 2022On Distillation of Guided Diffusion Modelscited by 301 here
  12. 2022DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Modelscited by 334 here
  13. 2022Scalable Diffusion Models with Transformerscited by 783 here
  14. 2023Consistency Modelscited by 477 here

The 24 most cited of 4,395.

Browse

30 of 2,909, ranked 6–35

  1. 6Elucidating the Design Space of Diffusion-Based Generative Models2022884
  2. 7Classifier-Free Diffusion Guidance2022812
  3. 8Scalable Diffusion Models with Transformers2022783
  4. 9Improved Denoising Diffusion Probabilistic Models2021772
  5. 10DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps2022679
  6. 11Scaling Rectified Flow Transformers for High-Resolution Image Synthesis2024480
  7. 12Consistency Models2023477
  8. 13SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis2023469
  9. 14Variational Diffusion Models2021355
  10. 15DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models2022334
All 30
  1. 16On Distillation of Guided Diffusion Models2022301
  2. 17Improved Techniques for Training Score-Based Generative Models2020283
  3. 18Sampling is as easy as learning the score: theory for diffusion models with minimal data assumptions2022282
  4. 19Pseudo Numerical Methods for Diffusion Models on Manifolds2022279
  5. 20Cascaded Diffusion Models for High Fidelity Image Generation2021279
  6. 21Maximum Likelihood Training of Score-Based Diffusion Models2021275
  7. 22Fast Sampling of Diffusion Models with Exponential Integrator2022274
  8. 23Structured Denoising Diffusion Models in Discrete State-Spaces2021243
  9. 24Score-based Generative Modeling in Latent Space2021241
  10. 25Progressive Distillation for Fast Sampling of Diffusion Models2022240
  11. 26PixArt-α: Fast Training of Diffusion Transformer for Photorealistic Text-to-Image Synthesis2023229
  12. 27Latent Consistency Models: Synthesizing High-Resolution Images with Few-Step Inference2023225
  13. 28One-step Diffusion with Distribution Matching Distillation2023220
  14. 29Diffusion Schr"odinger Bridge with Applications to Score-Based Generative Modeling2021202
  15. 30SiT: Exploring Flow and Diffusion-based Generative Models with Scalable Interpolant Transformers2024197
  16. 31Diffusion Models: A Comprehensive Survey of Methods and Applications2022194
  17. 32Adversarial Diffusion Distillation2023192
  18. 33Knowledge Distillation in Iterative Generative Models for Improved Sampling Speed2021184
  19. 34Convergence of denoising diffusion models under the manifold hypothesis2022182
  20. 35DeepCache: Accelerating Diffusion Models for Free2023179