{"id":"arxiv/2608.09870","url":"https://arxiv.org/abs/2608.09870","tags":["stat.ML","cs.LG","math.PR","math.ST"],"score":0,"title":"Logarithmic-Free Moment and Generalization Bounds for Uniformly Stable Algorithms","domain":"stat.ML","source":"arxiv","authors":["Thanh Nguyen-Cung","Binh T. Nguyen"],"reviews":0,"subject":"stat.ML","summary":"Uniform stability is a classical tool for controlling the generalization error of a learning algorithm. Bousquet, Klochkov, and Zhivotovskiy (2020) showed that the problem can be reduced to a moment inequality for a sum of weakly interacting functions of independent random variables. Their bound contains an additional factor $\\log n$, and they asked whether this factor can be removed. We answer this upper-bound question affirmatively. More specifically, let $Z=(Z_1,\\ldots,Z_n)$ have independent coordinates and let $g_i(Z)$ satisfy $\\mathbb E[g_i(Z)\\mid Z_{-i}]=0, \\ \\left| \\mathbb…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-08-10","reactions":{}}