mlsanityData to 7 Oct 2026

Sampling-Based Accuracy Testing of Posterior Estimators for General Inference

arXiv 2302.03026 ↗First posted 6 Feb 2023stat.ML astro-ph.IM cs.LGcited by 45

Not on the map. Why?

Abstract (arXiv)

Parameter inference, i.e. inferring the posterior distribution of the parameters of a statistical model given some data, is a central problem to many scientific disciplines. Generative models can be used as an alternative to Markov Chain Monte Carlo methods for conducting posterior inference, both in likelihood-based and simulation-based problems. However, assessing the accuracy of posteriors encoded in generative models is not straightforward. In this paper, we introduce `Tests of Accuracy with Random Points' (TARP) coverage testing as a method to estimate coverage probabilities of generative posterior estimators. Our method differs from previously-existing coverage-based methods, which require posterior evaluations. We prove that our approach is necessary and sufficient to show that a posterior estimator is accurate. We demonstrate the method on a variety of synthetic examples, and show that TARP can be used to test the results of posterior inference analyses in high-dimensional spaces. We also show that our method can detect inaccurate inferences in cases where existing methods fail.

Key facts. Cited by 45 papers in the map (data through 7 Oct 2026). Builds on 34 of its 74 references that are in the map. First posted on arXiv 6 Feb 2023.

Builds on

34 of 74 references.

References · 34

  1. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  2. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 8,964
  3. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,005
  4. 2017On Calibration of Modern Neural Networkscited by 4,860
  5. 2015Variational Inference with Normalizing Flowscited by 2,161
  6. 2019Normalizing Flows for Probabilistic Modeling and Inferencecited by 1,315
  7. 2014NICE: Non-linear Independent Components Estimationcited by 1,052
  8. 2019The frontier of simulation-based inferencecited by 490
  9. 2020Bayesian Deep Learning and a Probabilistic Perspective of Generalizationcited by 480
  10. 2018Sequential Neural Likelihood: Fast Likelihood-free Inference with Autoregressive Flowscited by 217
  11. 2019Automatic Posterior Transformation for Likelihood-Free Inferencecited by 216
  12. 2018Deep k-Nearest Neighbors: Towards Confident, Interpretable and Robust Deep Learningcited by 203
22 more
  1. 2018Bayesian Deep Convolutional Encoder-Decoder Networks for Surrogate Modeling and Uncertainty Quantificationcited by 199
  2. 2017Flexible statistical inference for mechanistic models of neural dynamicscited by 162
  3. 2021Benchmarking Simulation-Based Inferencecited by 160
  4. 2018Validating Bayesian Inference Algorithms with Simulation-Based Calibrationcited by 118
  5. 2020On Contrastive Learning for Likelihood-free Inferencecited by 96
  6. 2015Approximating Likelihood Ratios with Calibrated Discriminative Classifierscited by 87
  7. 2016Likelihood-free inference by ratio estimationcited by 80
  8. 2021Real-time gravitational-wave science with neural posterior estimationcited by 76
  9. 2020Neural Approximate Sufficient Statistics for Implicit Modelscited by 72
  10. 2018Likelihood-free inference with emulator networkscited by 61
  11. 2022Truncated proposals for scalable and hassle-free simulation-based inferencecited by 40
  12. 2022GATSBI: Generative Adversarial Training for Simulation-Based Inferencecited by 36
  13. 2019Conditional Density Estimation Tools in Python and R with Applications to Photometric Redshifts and Likelihood-Free Cosmological Inferencecited by 21
  14. 2022Posterior samples of source galaxies in strong gravitational lenses with score-based priorscited by 20
  15. 2019Mining for Dark Matter Substructure: Inferring subhalo population properties from strong lenses with machine learningcited by 15
  16. 2020Towards constraining warm dark matter with stellar streams through neural simulation-based inferencecited by 14
  17. 2021A neural simulation-based inference approach for characterizing the Galactic Center γ-ray excesscited by 12
  18. 2013Diagnostic tools of approximate Bayesian computation using the coverage propertycited by 10
  19. 2020Hierarchical Inference With Bayesian Neural Networks: An Application to Strong Gravitational Lensingcited by 10
  20. 2022Validation Diagnostics for SBI algorithms based on Normalizing Flowscited by 8
  21. 2021Simulation-based Bayesian inference for multi-fingered robotic graspingcited by 5
  22. 2020Simulation-Based Inference for Global Health Decisionscited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 8; 2024: 9; 2025: 12; 2026: 16.

45 papers.

Sort:

By area:

  1. 2025Simulation-Based Inference: A Practical Guidecited by 30
  2. 2024sbi reloaded: a toolkit for simulation-based inference workflowscited by 23
  3. 2025Effortless, Simulation-Efficient Bayesian Inference using Tabular Foundation Modelscited by 13
  4. 2023Balancing Simulation-based Inference for Conservative Posteriorscited by 10
  5. 2023Calibrating Neural Simulation-Based Inference with Differentiable Coverage Probabilitycited by 10
  6. 2025Multifidelity Simulation-based Inference for Computationally Expensive Simulatorscited by 9
  7. 2024PQMass: Probabilistic Assessment of the Quality of Generative Models using Probability Mass Estimationcited by 8
  8. 2024Conditional diffusions for amortized neural posterior estimationcited by 6
  9. 2023Variational Inference with Coverage Guarantees in Simulation-Based Inferencecited by 5
  10. 2024A Kernel-Based Conditional Two-Sample Test Using Nearest Neighbors (with Applications to Calibration, Regression Curves, and Simulation-Based Inference)cited by 5
  11. 2023Field-level simulation-based inference with galaxy catalogs: the impact of systematic effectscited by 4
  12. 2023Simulation-based Inference with the Generalized Kullback-Leibler Divergencecited by 3
33 more
  1. 2023Bayesian Imaging for Radio Interferometry with Score-Based Priorscited by 3
  2. 2023Discriminative calibration: Check Bayesian computation from simulations and flexible classifiercited by 2
  3. 2024Statistical Patterns in the Equations of Physics and the Emergence of a Meta-Law of Naturecited by 2
  4. 2024Full-waveform earthquake source inversion using simulation-based inferencecited by 2
  5. 2025Blind Strong Gravitational Lensing Inversion: Joint Inference of Source and Lens Mass with Score-Based Modelscited by 2
  6. 2026Training-Free Bayesian Filtering with Generative Emulatorscited by 2
  7. 2026Generative Models and Statistical Validationcited by 2
  8. 2024Simulation-based inference with scattering representations: scattering is all you needcited by 1
  9. 2025Conformal C2ST: Turning weak classifiers into strong two-sample testscited by 1
  10. 2025Trustworthy scientific inference with generative modelscited by 1
  11. 2025On Uncertainty Calibration for Equivariant Functionscited by 1
  12. 2026First Estimation of Model Parameters for Neutrino-Induced Nucleon Knockout Using Simulation-Based Inferencecited by 1
  13. 2026MIRA: A Score for Conditional Distribution Accuracy and Model Comparisoncited by 1
  14. 2026Pointwise Metrics Mislead: An Evaluation Protocol for Multimodal Inverse Problemscited by 1
  15. 2026An Introduction to Bayesian and Frequentist Simulation-Based Inference with Machine Learningcited by 1
  16. 2023Exposing Disparities in Flood Adaptation for Equitable Future Interventionscited by 0
  17. 2024Hybrid summary statistics: neural weak lensing inference beyond the power spectrumcited by 0
  18. 2024Noise-Aware Differentially Private Variational Inferencecited by 0
  19. 2025CoLT: The conditional localization test for assessing the accuracy of neural posterior estimatescited by 0
  20. 2025Set-based Implicit Likelihood Inference of Galaxy Cluster Masscited by 0
  21. 2025Noise-Guided Transport for Imitation Learningcited by 0
  22. 2025Bridging Simulators with Conditional Optimal Transportcited by 0
  23. 2025Dark Energy Survey Year 3 results: Simulation-based wCDM inference from weak lensing and galaxy clustering maps with deep learning: Analysis designcited by 0
  24. 2026Inverse Problems Conditioned on Observation Ensembles: Applications and Methodscited by 0
  25. 2026A Fast Generative Framework for High-dimensional Posterior Sampling: Application to CMB Delensingcited by 0
  26. 2026Improving moment tensor solutions under Earth structure uncertainty with simulation-based inferencecited by 0
  27. 2026Tokenised Flow Matching for Hierarchical Simulation Based Inferencecited by 0
  28. 2026Pre-trained Tabular Foundation Models as Versatile Summary Networks for Neural Posterior Estimationcited by 0
  29. 2026Conservative neural posterior estimation via distributionally robust trainingcited by 0
  30. 2026Learning the Universe: Posterior Reliability of Neural Generative Models in High-Dimensional Field-Level Inference of Cosmic Initial Conditionscited by 0
  31. 2026The Degeneracy Distillerycited by 0
  32. 2026Posterior Inference of Hamiltonian Parameters from RIXS Spectroscopycited by 0
  33. 2026Simulation-Based Inference for Plate Reverb System Identificationcited by 0