Identifiability and Stability of Generative Drifting with Companion-Elliptic Kernel Families

arXiv:2604.24196v1 Announce Type: cross Abstract: This paper analyzes identifiability and stability for the drifting field underlying distributional matching in the Generative Drifting framework of Deng et al. First, we introduce the class of companion-elliptic kernels, which includes the Laplace kernel and is characterized by a second-order elliptic coupling between each kernel kappa in this class and its companion function eta. For each kernel in this class and each pair of Borel probability measures, we prove that the drifting field vanishes if and only if the two probability measures are equal. We further show that this class consists precisely of Gaussian kernels and Matern kernels with nu ge 1/2. Second, by constructing counterexamples, we exhibit sequences for which mass escapes to infinity while the field tends to zero; in particular, control of the field norm alone does not guarantee weak convergence. Nevertheless, we prove that the only possible mode of failure is confined to the one-dimensional ray {c,p:0le cle 1}. Consequently, weak convergence can be restored by imposing an asymptotic lower bound on the intrinsic overlap scalar, a linear observable defined by the kernel and the target measure.

Related

• Branching Flows: Discrete, Continuous, and Manifold Flow Matching with Splits and Deletions
• A Divergence-Based Method for Weighting and Averaging Model Predictions
• Identifiability of Potentially Degenerate Gaussian Mixture Models With Piecewise Affine Mixing
• Sliced Wasserstein Steering between Gaussian Measures

Source: arXiv cs.LG | 2026-04-28