{"id":"arxiv/2609.33629","url":"https://arxiv.org/abs/2609.33629","tags":["math.NA"],"score":0,"title":"Wasserstein Bounds for Ambient and Latent Smoothing on Data Manifolds","domain":"math.NA","source":"arxiv","authors":["Wonjun Lee","Wenyan Luo"],"reviews":0,"subject":"math.NA","summary":"Smoothing empirical data can fill gaps between observations, while noise in the surrounding space can move mass away from the manifold supporting the data. We derive Wasserstein error bounds for ambient Gaussian smoothing and for smoothing through an encoder--decoder representation. For a compact connected smooth manifold without boundary and a smooth positive target density, established heat-flow and entropy estimates imply that sufficiently small intrinsic heat smoothing improves the squared intrinsic Wasserstein error. Coupling this estimate with ambient noise yields a bound whose…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-09-27","reactions":{}}