Finite-Probe Total-Variation Certificates for Finite-Basis Drifting Models
Sam Andersson, Ricky Molén
stat.ML, cs.IT, cs.LG
Submitted: 2026-08-03
Comments: 39 pages, 8 figures. Reproducibility code and numerical outputs: https://github.com/SamAndersson-C/finite-drift-certification
Code: https://github.com/SamAndersson-C/finite-drift-certification
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Finite-Particle Convergence Rates for Conservative and Non-Conservative Drifting Models
- Algebraic compressed sensing
- Gradient Flow Drifting: Generative Modeling via Wasserstein Gradient Flows of KDE-Approximated Divergences
- Generative Modeling via Drifting
- Kernel-Gradient Drifting Models
- DriftXpress: Faster Drifting Models via Projected RKHS Fields
- Drifting Fields are not Conservative
- A Kernel Method for the Two-Sample Problem
- On the Wasserstein Gradient Flow Interpretation of Drifting Models
- Sinkhorn-Drifting Generative Models
- A Unified View of Score-Based and Drifting Models
- Identifiability and Stability of Generative Drifting in the Companion-Elliptic Kernel Family
- Kernel Tests of Equivalence
- Empirical Bernstein Bounds and Sample Variance Penalization
- The Zero Set of a Real Analytic Function
- Generative Drifting is Secretly Score Matching: a Spectral and Variational Perspective
- The Score-Difference Flow for Implicit Generative Modeling
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
- Scalable Posterior Uncertainty for Flexible Density-Based Clustering
- A Transport Geometry of Benign Overfitting
- Efficient quadratic entropy with distance sketches