Score-Based Generative Modeling

Overview

Score-based generative modeling learns the gradient of a data distribution’s log density, often at multiple noise levels using denoising score matching. Generation then follows learned scores through reverse-time stochastic differential equations, probability-flow ordinary differential equations, or Langevin correctors; extensions handle manifolds, discrete data, and scientific inverse problems. Methods are evaluated on image-generation benchmarks and tasks such as MRI reconstruction, simulation-based inference, and data assimilation, where fidelity, calibration, and sampling cost all matter. A central challenge is connecting score-estimation accuracy to reliable samples: low-noise errors, discretization, initialization, and data geometry can all affect results, while stronger smoothing may reduce memorization but alter fidelity and diversity.

Latest papers

CardsList

122 papers10 papers in the last four weeks, down 41% on the four weeks before. 0.1% of all new papers.

  1. Amortizing Tempered Sampling with Corrected Self-Training

    Nov 17, 2026Hoang-Viet Nguyen, Hai-Dang Dau, Alexandre H. ThiérySampling from Unnormalized DistributionsDiffusion Model Sampling

  2. Diffusion-free Score Matching for Exact Conditional Sampling

    Nov 6, 2026Thomas Wedenig, Robert PeharzScore-Based Generative ModelingScore Matching

  3. Generative diffusion in long-range correlated landscapes: Speciation cascade and the foam structure of the score

    Nov 2, 2026Jerome Garnier-Brun, Giulio Biroli, Marc Mézard +1Diffusion SamplingDiffusion Models

  4. Finite-Time Online Sample Complexity for Conditional Diffusion Models

    Nov 1, 2026Mudit Gaur, Prashant Trivedi, Sasidhar Kunapuli +2Conditional Diffusion ModelsDiffusion Models

  5. Image Modeling: From Handcrafted Priors to Foundation Image Models

    Oct 27, 2026A. Murat TekalpImage Inverse ProblemsConditional Generative Modeling

  6. Approximation Families and the Statistical Efficiency of Denoising Score Matching

    Oct 26, 2026Brian MiyatakeScore MatchingScore-Based Generative Modeling

  7. Generative AI in Turbulence: Foundations, Applications, Validations and Benchmarks

    Oct 26, 2026Luca Biferale, Michele BuzzicottiGenerative ModelingGenerative AI Evaluation

  8. OpenMAP-HIE: Pathology Informed Score-based Diffusion Model for Unsupervised Neonatal Hypoxic-Ischemic Encephalopathy Lesion Segmentation

    Oct 21, 2026Pengyu Kan, Craig Jones, Kenichi OishiMedical Image SegmentationMedical Image Anomaly Detection

  9. Diffusion Score Estimation with Shallow Neural Networks: Regularity, Approximation, and Finite-Sample Rates

    Oct 19, 2026Fan Liu, Yahong Yang, Wei ZhuScore-Based Generative ModelingDiffusion Models

  10. Two-Stage Score Estimation for Group-Specific Diffusion Models

    Oct 14, 2026Huitao Yang, Oscar Hernan Madrid PadillaDiffusion ModelsScore-Based Generative Modeling

  11. Denoising Score Matching in Diffusion Models: The Effect of Rank Constraints on Generalization

    Oct 14, 2026Anand Jerry George, Nicolas MacrisScore-Based Generative ModelingNeural Network Generalization

  12. Colored Graph Signal Diffusion Models

    Oct 12, 2026Yiğit Berkay Uslu, Samar Hadou, Shirin Saeedi Bidokhti +1Diffusion ModelsGraph Diffusion

  13. A Stochastic Analysis Perspective on Diffusion Models

    Oct 9, 2026Stanislas StrasmanScore-Based Generative ModelingDiffusion Sampling

  14. Diffusion Removes Langevin's Conditioning Dependence: A Sharp Gaussian Analysis

    Oct 8, 2026Adam Perbost, Francis Bach, Pierre MarionDiffusion SamplingScore-Based Generative Modeling

  15. A Unified Score Matching Paradigm for Video Anomaly Detection and Anticipation

    Oct 8, 2026Congqi Cao, Zhenhe Liang, Hanwen Zhang +4Video Anomaly DetectionScore-Based Generative Modeling

  16. Gradient-Free Sampling from Generative Models via Stochastic Bounded Extremum Seeking

    Oct 3, 2026Alexander ScheinkerDiffusion SamplingMarkov Chain Monte Carlo

  17. Sample complexity bounds for categorical Markov random fields via Discrete Diffusions

    Oct 1, 2026Shivam Kumar, Nabarun DebMarkov Random FieldsDiffusion Sampling

  18. Ornstein-Uhlenbeck Is Hard to Beat, Yet Superlinear Drift Ships Lower Transport Costs

    Sep 29, 2026Attila Lovas, Lóránt NagyDiffusion-Based Image GenerationScore-Based Generative Modeling

  19. Weighting Schedules Govern What and When Score-Based Generative Models Learn from Multimodal Data

    Sep 28, 2026Jérémie Klinger, Raphaël Urfin, Giulio Biroli +1Generative ModelingScore-Based Generative Modeling

  20. AECSF: Adaptive Ensemble Conditional Score Filtering for High-Dimensional Nonlinear Data Assimilation

    Sep 26, 2026Yangwen Zhang, Shiwei Ni, Xiaoping Zhang +2Data AssimilationBayesian Filtering

  21. Quantum score matching with applications to learning thermal states

    Sep 23, 2026Yulong Dong, Jiaqi LengQuantum Machine LearningScore-Based Generative Modeling

  22. Robustness of Diffusion Models under Distribution Shift

    Sep 23, 2026Wei Luo, Neil K. Chada, Shijie Zhang +1Distribution Shift RobustnessDiffusion Sampling

  23. Walking the Score Manifold: Continuous-time Generative Dynamics on Learned Data Manifolds

    Sep 15, 2026Jan Tauberschmidt, Brian B. Moser, Stanislav Frolov +3Dynamical SystemsLatent Dynamics Modeling

  24. Score-based Outlier Generation via Controlling the Radon-Nikodym Derivative

    Sep 10, 2026Amartya Mukherjee, Tristan Milne, Kry Yik-Chau Lui +2Diffusion SamplingSynthetic Anomaly Generation