{"id":"arxiv/2607.00665","url":"https://arxiv.org/abs/2607.00665","tags":["math.OC"],"score":0,"title":"Effective dynamics of the Sinkhorn algorithm in the regime of low entropy regularization","domain":"math.OC","source":"arxiv","authors":["Guillaume Wang"],"reviews":0,"subject":"math.OC","summary":"The Sinkhorn algorithm is the de facto standard method for numerically solving entropy-regularized optimal transport problems over finite sets. In this work, we investigate a phenomenon arising when Sinkhorn is applied with a small regularization parameter $τ$: the evolution of the dual variables (the logarithm of the scaling factors) is approximately piecewise-linear, while the primal variables (the approximate transport plans) exhibit a saddle-to-saddle type behavior. We prove that as $τ\\to 0$, the Sinkhorn iterates indeed converge to a continuous-time curve consistent with these…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-07-01","reactions":{}}