Two-Sample Testing via Generative Processes
stat.ML, cs.LG
Submitted: 2026-10-06
Updated: 2026-10-06
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Sources
- On the Closed-Form of Flow Matching: Generalization Does Not Arise from Target Stochasticity
- Density Ratio Estimation via Infinitesimal Classification
- Text-to-Image Diffusion Models are Zero-Shot Classifiers
- Two-Sample Testing via Path-based Inference
- AutoGluon-Tabular: Robust and Accurate AutoML for Structured Data
- Denoising diffusion models for out-of-distribution detection
- Mutual Information and Minimum Mean-square Error in Gaussian Channels
- Classification accuracy as a proxy for two sample testing
- Minimax optimality of permutation tests
- Information-Theoretic Diffusion
- A Witness Two-Sample Test
- Your Diffusion Model is Secretly a Zero-Shot Classifier
- Unsupervised Out-of-Distribution Detection with Diffusion Inpainting
- Interpretation and Generalization of Score Matching
- Diffusion-Based Hypothesis Testing and Change-Point Detection
- Midpoint Generative Models
- Zero-Flow Two-Sample Tests
Related papers
- Behavior of prediction performance metrics with rare events
- Optimal Estimation of Generic Dynamics by Path-Dependent Neural Jump ODEs
- A Posterior-Dynamics Framework for Imaging Inverse Problems with Pretrained Diffusion Priors
- Efficient quadratic entropy with distance sketches
- Embedding-Bias in Conditional Independence Testing
- From Log-Odds to Shapley Values: An Explanatory Geometry for the Weighted Naive Bayes Classifier