{"id":"arxiv/2607.15505","url":"https://arxiv.org/abs/2607.15505","tags":["math.OC","cs.LG","cs.NE","math.NA"],"score":0,"title":"Fast and Scalable Caputo Fractional Gradient Descent via Perturbation-Preserving Memory Compression","domain":"math.OC","source":"arxiv","authors":["Hwanseo Lee","Junseo Lee","Hyunju Kim"],"reviews":0,"subject":"math.OC","summary":"Fractional gradient descent (FGD) incorporates long-range memory through Caputo-type operators and has been shown to improve stability in ill-conditioned and nonconvex optimization problems. Despite these advantages, its practical use remains limited, mainly due to the high computational cost of evaluating history-dependent convolutions, which scales quadratically with the number of iterations. In this paper, we focus on making Caputo-based optimization computationally viable without sacrificing its intrinsic memory structure. We begin by expressing the fractional descent direction as a…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-07-16","reactions":{}}