mlsanityData to 14 Oct 2026

Probability Density from Latent Diffusion Models for Out-of-Distribution Detection

arXiv 2508.15737 ↗First posted 21 Aug 2025cs.LG cs.CVcited by 2

On the map: Diffusion models → Medical imaging and segmentation

Abstract (arXiv)

Despite rapid advances in AI, safety remains the main bottleneck to deploying machine-learning systems. A critical safety component is out-of-distribution detection: given an input, decide whether it comes from the same distribution as the training data. In generative models, the most natural OOD score is the data likelihood. Actually, under the assumption of uniformly distributed OOD data, the likelihood is even the optimal OOD detector, as we show in this work. However, earlier work reported that likelihood often fails in practice, raising doubts about its usefulness. We explore whether, in practice, the representation space also suffers from the inability to learn good density estimation for OOD detection, or if it is merely a problem of the pixel space typically used in generative models. To test this, we trained a Variational Diffusion Model not on images, but on the representation space of a pre-trained ResNet-18 to assess the performance of our likelihood-based detector in comparison to state-of-the-art methods from the OpenOOD suite.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Placed in Medical imaging and segmentation, one of 14 directions in Diffusion models. Builds on 33 of its 58 references that are in the map. First posted on arXiv 21 Aug 2025.

Builds on

33 of 58 references.

Closest ancestors · 2

Same direction, newest first

  1. 2023Likelihood-based Out-of-Distribution Detection with Denoising Diffusion Probabilistic Models26 Oct
  2. 2022Denoising diffusion models for out-of-distribution detection14 Nov

Other references · 31

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  3. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  4. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 9,122
  5. 2022Classifier-Free Diffusion Guidancecited by 5,847
  6. 2016Layer Normalizationcited by 5,669
  7. 2017On Calibration of Modern Neural Networkscited by 4,960
  8. 2016Gaussian Error Linear Units (GELUs)cited by 3,502
  9. 2013Describing Textures in the Wildcited by 2,125
  10. 2018A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attackscited by 1,485
  11. 2020Energy-based Out-of-distribution Detectioncited by 1,299
  12. 2021Variational Diffusion Modelscited by 1,153
19 more
  1. 2018Deep Anomaly Detection with Outlier Exposurecited by 942
  2. 2021Generalized Out-of-Distribution Detection: A Surveycited by 802
  3. 2014Deep Neural Networks are Easily Fooled: High Confidence Predictions for Unrecognizable Imagescited by 776
  4. 2022Out-of-Distribution Detection with Deep Nearest Neighborscited by 600
  5. 2019Scaling Out-of-Distribution Detection for Real-World Settingscited by 465
  6. 2018Do Deep Generative Models Know What They Don't Know?cited by 443
  7. 2022ViM: Out-Of-Distribution with Virtual-logit Matchingcited by 405
  8. 2020Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awarenesscited by 389
  9. 2019Sliced Score Matching: A Scalable Approach to Density and Score Estimationcited by 384
  10. 2021Open-Set Recognition: a Good Closed-Set Classifier is All You Need?cited by 353
  11. 2020Why Normalizing Flows Fail to Detect Out-of-Distribution Datacited by 221
  12. 2022Extremely Simple Activation Shaping for Out-of-Distribution Detectioncited by 212
  13. 2021A Simple Fix to Mahalanobis Distance for Improving Near-OOD Detectioncited by 207
  14. 2021Deep Deterministic Uncertainty: A Simple Baselinecited by 200
  15. 2023OpenOOD v1.5: Enhanced Benchmark for Out-of-Distribution Detectioncited by 179
  16. 2022Is Out-of-Distribution Detection Learnable?cited by 109
  17. 2019Detecting semantic anomaliescited by 50
  18. 2024Feature Density Estimation for Out-of-Distribution Detection via Normalizing Flowscited by 4
  19. 2024NODI: Out-Of-Distribution Detection with Noise from Diffusioncited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026Backbone-Equated Diffusion OOD via Sparse Internal Snapshotscited by 0
  2. 2026Tippett-minimum Fusion of Representation-space Diffusion Models for Multi-Encoder Out-of-Distribution Detectioncited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Revisiting Likelihood-Based Out-of-Distribution Detection by Modeling Representations
  2. 2024A Geometric Explanation of the Likelihood OOD Detection Paradox
  3. 2026Geometry over Density: Few-Shot Cross-Domain OOD Detection
  4. 2024Approximations to the Fisher Information Metric of Deep Generative Models for Out-Of-Distribution Detection
  5. 2024Enhancing OOD Detection Using Latent Diffusion
  6. 2024Exploiting Diffusion Prior for Out-of-Distribution Detection
  7. 2025Beyond Binary Out-of-Distribution Detection: Characterizing Distributional Shifts with Multi-Statistic Diffusion Trajectories
  8. 2025EigenScore: OOD Detection using Covariance in Diffusion Models