{"id":"arxiv/2602.05174","url":"https://arxiv.org/abs/2602.05174","tags":["stat.ML","cs.AI","cs.LG","math.ST"],"score":0,"title":"Total Variation Rates for Riemannian Flow Matching","domain":"stat.ML","source":"arxiv","authors":["Yunrui Guan","Krishnakumar Balasubramanian","Shiqian Ma"],"reviews":0,"subject":"stat.ML","summary":"Riemannian flow matching (RFM) extends flow-based generative modeling to data supported on manifolds by learning a time-dependent tangent vector field whose flow-ODE transports a simple base distribution to the data law. We develop a nonasymptotic Total Variation (TV) convergence analysis for RFM samplers that use a learned vector field together with Euler discretization on manifolds. Our key technical ingredient is a differential inequality governing the evolution of TV between two manifold ODE flows, which expresses the time-derivative of TV through the divergence of the vector-field…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-02-05","reactions":{}}