Uniform-in-time propagation-of-chaos bounds for SVGD are obtained via cutoff for distributional metrics (logarithmic rates) and via finite-dimensional closure plus conjugacy for Gaussian targets (parametric N^{-1/2} rates).
On the rate of convergence in wasserstein distance of the empirical measure.Probability theory and related fields, 162(3):707–738
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3representative citing papers
An explicit nearly-optimal data-dependent DKW inequality yields uniform 1-δ confidence bands for the CDF of a functional of a regenerative Markov chain from its empirical CDF.
Span-level Wasserstein distances between hidden-state distributions of correct and incorrect rollouts provide a self-supervised signal to reweight advantages in GRPO, improving fine-grained credit assignment on math and code tasks.
citing papers explorer
-
Uniform-in-time Propagation-of-Chaos for Stein Variational Gradient Descent
Uniform-in-time propagation-of-chaos bounds for SVGD are obtained via cutoff for distributional metrics (logarithmic rates) and via finite-dimensional closure plus conjugacy for Gaussian targets (parametric N^{-1/2} rates).
-
A data-dependent DKW inequality for regenerative Markov chains
An explicit nearly-optimal data-dependent DKW inequality yields uniform 1-δ confidence bands for the CDF of a functional of a regenerative Markov chain from its empirical CDF.
-
Hidden States Know Where Reasoning Diverges: Credit Assignment via Span-Level Wasserstein Distance
Span-level Wasserstein distances between hidden-state distributions of correct and incorrect rollouts provide a self-supervised signal to reweight advantages in GRPO, improving fine-grained credit assignment on math and code tasks.