mlsanityData to 7 Oct 2026

Model-agnostic out-of-distribution detection using combined statistical tests

arXiv 2203.01097 ↗First posted 2 Mar 2022stat.ML cs.LGcited by 14

Not on the map. Why?

Abstract (arXiv)

We present simple methods for out-of-distribution detection using a trained generative model. These techniques, based on classical statistical tests, are model-agnostic in the sense that they can be applied to any differentiable generative model. The idea is to combine a classical parametric test (Rao's score test) with the recently introduced typicality test. These two test statistics are both theoretically well-founded and exploit different sources of information based on the likelihood for the typicality test and its gradient for the score test. We show that combining them using Fisher's method overall leads to a more accurate out-of-distribution test. We also discuss the benefits of casting out-of-distribution detection as a statistical testing problem, noting in particular that false positive rate control can be valuable for practical out-of-distribution detection. Despite their simplicity and generality, these methods can be competitive with model-specific out-of-distribution detection algorithms without any assumptions on the out-distribution.

Key facts. Cited by 14 papers in the map (data through 7 Oct 2026). Builds on 32 of its 70 references that are in the map. First posted on arXiv 2 Mar 2022.

Builds on

32 of 70 references.

References · 32

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2016Overcoming catastrophic forgetting in neural networkscited by 4,718
  3. 2017Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithmscited by 4,283
  4. 2014Deep Learning Face Attributes in the Wildcited by 3,679
  5. 2018Glow: Generative Flow with Invertible 1x1 Convolutionscited by 1,539
  6. 2018A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attackscited by 1,472
  7. 2018Deep Anomaly Detection with Outlier Exposurecited by 939
  8. 2015Optimizing Neural Networks with Kronecker-factored Approximate Curvaturecited by 776
  9. 2016Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networkscited by 541
  10. 2015Importance Weighted Autoencoderscited by 454
  11. 2019Likelihood Ratios for Out-of-Distribution Detectioncited by 451
  12. 2018Do Deep Generative Models Know What They Don't Know?cited by 443
20 more
  1. 2014New insights and perspectives on the natural gradient methodcited by 435
  2. 2020Why Normalizing Flows Fail to Detect Out-of-Distribution Datacited by 220
  3. 2019Input complexity and out-of-distribution detection with likelihood-based generative modelscited by 186
  4. 2020Contrastive Training for Improved Out-of-Distribution Detectioncited by 186
  5. 2018Handling Incomplete Heterogeneous Data using VAEscited by 148
  6. 2020Likelihood Regret: An Out-of-Distribution Detection Score For Variational Auto-encodercited by 145
  7. 2018Inference Suboptimality in Variational Autoencoderscited by 131
  8. 2020Learning Deep Kernels for Non-Parametric Two-Sample Testscited by 129
  9. 2019BIVA: A Very Deep Hierarchy of Latent Variables for Generative Modelingcited by 109
  10. 2021Understanding Failures in Out-of-Distribution Detection with Deep Generative Modelscited by 89
  11. 2018EDDI: Efficient Dynamic Discovery of High-Value Information with Partial VAEcited by 85
  12. 2019BackPACK: Packing more into backpropcited by 77
  13. 2020Density of States Estimation for Out-of-Distribution Detectioncited by 74
  14. 2020Understanding Anomaly Detection with Deep Invertible Networks through Hierarchies of Distributions and Featurescited by 64
  15. 2020not-MIWAE: Deep Generative Modelling with Missing not at Random Datacited by 61
  16. 2021Hierarchical VAEs Know What They Don't Knowcited by 58
  17. 2015Efficient Per-Example Gradient Computationscited by 42
  18. 2019Detecting Out-of-Distribution Inputs to Deep Generative Models Using Typicalitycited by 30
  19. 2019Efficient Per-Example Gradient Computations in Convolutional Neural Networkscited by 17
  20. 2021Robust Out-of-Distribution Detection on Deep Probabilistic Generative Modelscited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 5; 2023: 1; 2024: 7; 2025: 0; 2026: 1.

14 papers.

Sort:

By area:

  1. 2023Unsupervised Out-of-Distribution Detection with Diffusion Inpaintingcited by 57
  2. 2022Boosting Out-of-distribution Detection with Typical Featurescited by 48
  3. 2022Out-of-Distribution Detection with Reconstruction Error and Typicality-based Penaltycited by 11
  4. 2022Boosting Out-of-Distribution Detection with Multiple Pre-trained Modelscited by 7
  5. 2024Diffusion-based Layer-wise Semantic Reconstruction for Unsupervised Out-of-Distribution Detectioncited by 5
  6. 2024A Unified Data Representation Learning for Non-parametric Two-sample Testingcited by 5
  7. 2024Distribution Guided Active Feature Acquisitioncited by 3
  8. 2022Multiple Testing Framework for Out-of-Distribution Detectioncited by 2
  9. 2024Approximations to the Fisher Information Metric of Deep Generative Models for Out-Of-Distribution Detectioncited by 2
  10. 2022Out-of-Distribution Detection using BiGAN and MDLcited by 1
  11. 2024Thinking in Groups: Permutation Tests Reveal Near-Out-of-Distributioncited by 1
  12. 2026An Isotropic Approach to Efficient Uncertainty Quantification with Gradient Normscited by 1
2 more
  1. 2024Rethinking Test-time Likelihood: The Likelihood Path Principle and Its Application to OOD Detectioncited by 0
  2. 2024DSDE: Using Proportion Estimation to Improve Model Selection for Out-of-Distribution Detectioncited by 0