{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,5]],"date-time":"2026-07-05T04:29:49Z","timestamp":1783225789645,"version":"3.54.6"},"reference-count":66,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"9","license":[{"start":{"date-parts":[[2025,9,1]],"date-time":"2025-09-01T00:00:00Z","timestamp":1756684800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2025,9,1]],"date-time":"2025-09-01T00:00:00Z","timestamp":1756684800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,9,1]],"date-time":"2025-09-01T00:00:00Z","timestamp":1756684800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"name":"Singapore Ministry of Education AcRF Tier 2","award":["A-8000423-00-00"],"award-info":[{"award-number":["A-8000423-00-00"]}]},{"name":"AcRF Tier 1","award":["A-8000980-00-00"],"award-info":[{"award-number":["A-8000980-00-00"]}]},{"name":"AcRF Tier 1","award":["A-8002934-00-00"],"award-info":[{"award-number":["A-8002934-00-00"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Inform. Theory"],"published-print":{"date-parts":[[2025,9]]},"DOI":"10.1109\/tit.2025.3588401","type":"journal-article","created":{"date-parts":[[2025,7,14]],"date-time":"2025-07-14T17:42:18Z","timestamp":1752514938000},"page":"7270-7291","source":"Crossref","is-referenced-by-count":4,"title":["On the Convergence of (Stochastic) Gradient Descent for Kolmogorov\u2013Arnold Networks"],"prefix":"10.1109","volume":"71","author":[{"given":"Yihang","family":"Gao","sequence":"first","affiliation":[{"name":"Department of Mathematics, National University of Singapore, Queenstown, Singapore"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5008-4527","authenticated-orcid":false,"given":"Vincent Y. F.","family":"Tan","sequence":"additional","affiliation":[{"name":"Department of Mathematics, Department of Electrical and Computer Engineering, National University of Singapore, Queenstown, Singapore"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1090\/trans2\/017\/12"},{"issue":"5","key":"ref2","first-page":"953","article-title":"On the representation of continuous functions of many variables by superposition of continuous functions of one variable and addition","volume":"114","year":"1957","journal-title":"Doklady Akademii Nauk"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1007\/s00365-009-9054-2"},{"key":"ref4","first-page":"1","article-title":"KAN: Kolmogorov\u2013Arnold networks","volume-title":"Proc. 13th Int. Conf. Learn. Represent.","author":"Liu"},{"key":"ref5","article-title":"Convolutional Kolmogorov\u2013Arnold networks","author":"Dylan Bodner","year":"2024","journal-title":"arXiv:2406.13155"},{"key":"ref6","article-title":"Kolmogorov\u2013Arnold convolutions: Design principles and empirical studies","author":"Drokin","year":"2024","journal-title":"arXiv:2407.01092"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v39i5.32491"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.2139\/ssrn.4825654"},{"key":"ref9","article-title":"A temporal Kolmogorov\u2013Arnold transformer for time series forecasting","author":"Genet","year":"2024","journal-title":"arXiv:2406.02486"},{"key":"ref10","article-title":"Kolmogorov\u2013Arnold networks for time series: Bridging predictive power and interpretability","author":"Xu","year":"2024","journal-title":"arXiv:2406.02496"},{"key":"ref11","article-title":"Are KANs effective for multivariate time series forecasting?","author":"Han","year":"2024","journal-title":"arXiv:2408.11306"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.23919\/ICCAS63016.2024.10773080"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP49660.2025.10888327"},{"key":"ref14","article-title":"GraphKAN: Enhancing feature extraction with graph Kolmogorov Arnold networks","author":"Zhang","year":"2024","journal-title":"arXiv:2406.13597"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1016\/j.cma.2024.117518"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1007\/s00366-025-02160-w"},{"key":"ref17","article-title":"Inferring turbulent velocity and temperature fields and their statistics from Lagrangian velocity measurements using physics-informed Kolmogorov\u2013Arnold networks","author":"Diego Toscano","year":"2024","journal-title":"arXiv:2407.15727"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1016\/j.cma.2024.117699"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2024.3504962"},{"key":"ref20","article-title":"Finite basis Kolmogorov\u2013Arnold networks: Domain decomposition for data-driven and physics-informed problems","author":"Howard","year":"2024","journal-title":"arXiv:2406.19662"},{"key":"ref21","article-title":"Kolmogorov\u2013Arnold networks in molecular dynamics","author":"Nagai","year":"2024","journal-title":"arXiv:2407.17774"},{"key":"ref22","article-title":"Demonstrating the efficacy of Kolmogorov\u2013Arnold networks in vision tasks","author":"Cheon","year":"2024","journal-title":"arXiv:2406.14916"},{"key":"ref23","article-title":"CoxKAN: Kolmogorov\u2013Arnold networks for interpretable, high-performance survival analysis","author":"Knottenbelt","year":"2024","journal-title":"arXiv:2409.04290"},{"key":"ref24","article-title":"ReLU-KAN: New Kolmogorov\u2013Arnold networks that only need matrix addition, dot multiplication, and ReLU","author":"Qiu","year":"2024","journal-title":"arXiv:2406.02075"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1016\/j.engappai.2020.104137"},{"key":"ref26","article-title":"Gaussian process Kolmogorov\u2013Arnold networks","author":"Siyuan Chen","year":"2024","journal-title":"arXiv:2407.18397"},{"key":"ref27","article-title":"A comprehensive survey on Kolmogorov Arnold networks (KAN)","author":"Ji","year":"2024","journal-title":"arXiv:2407.11075"},{"key":"ref28","article-title":"KAN or MLP: A fairer comparison","author":"Yu","year":"2024","journal-title":"arXiv:2407.16674"},{"key":"ref29","article-title":"KAN 2.0: Kolmogorov\u2013Arnold networks meet science","author":"Liu","year":"2024","journal-title":"arXiv:2408.10205"},{"key":"ref30","article-title":"Chebyshev polynomial-based Kolmogorov\u2013Arnold networks: An efficient architecture for nonlinear function approximation","author":"Sidharth","year":"2024","journal-title":"arXiv:2405.07200"},{"key":"ref31","article-title":"RKAN: Rational Kolmogorov\u2013Arnold networks","author":"Afzal Aghaei","year":"2024","journal-title":"arXiv:2406.14495"},{"key":"ref32","first-page":"1","article-title":"ADAM: A method for stochastic optimization","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Kingma"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1137\/17M1114296"},{"key":"ref34","first-page":"983","article-title":"On the convergence of stochastic gradient descent with adaptive stepsizes","volume-title":"Proc. 22nd Int. Conf. Artif. Intell. Statist.","author":"Li"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1007\/s10107-006-0706-8"},{"issue":"7","key":"ref36","first-page":"2121","article-title":"Adaptive subgradient methods for online learning and stochastic optimization","volume":"12","author":"Duchi","year":"2011","journal-title":"J. Mach. Learn. Res."},{"key":"ref37","first-page":"1","article-title":"On the convergence of Adam and beyond","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Reddi"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1145\/3446776"},{"key":"ref39","article-title":"No bad local minima: Data independent training error guarantees for multilayer neural networks","author":"Soudry","year":"2016","journal-title":"arXiv:1605.08361"},{"key":"ref40","first-page":"1","article-title":"Gradient descent provably optimizes over-parameterized neural networks","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Du"},{"key":"ref41","first-page":"10676","article-title":"Gradient descent finds the global optima of two-layer physics-informed neural networks","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Gao"},{"key":"ref42","first-page":"1","article-title":"Learning ReLUs via gradient descent","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"30","author":"Soltanolkotabi"},{"key":"ref43","first-page":"1216","article-title":"Diverse neural network learns true target functions","volume-title":"Proc. 20th Int. Conf. Artif. Intell. Statist.","author":"Xie"},{"key":"ref44","first-page":"1","article-title":"On the global convergence of gradient descent for over-parameterized models using optimal transport","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"31","author":"Chizat"},{"key":"ref45","first-page":"1","article-title":"Neural tangent kernel: Convergence and generalization in neural networks","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"31","author":"Jacot"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1109\/TIT.2018.2854560"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1016\/j.jcp.2021.110768"},{"key":"ref48","first-page":"1195","article-title":"Fast and faster convergence of SGD for over-parameterized models and an accelerated perceptron","volume-title":"Proc. 22nd Int. Conf. Artif. Intell. Statist.","author":"Vaswani"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1109\/JSAIT.2020.2991332"},{"key":"ref50","first-page":"242","article-title":"A convergence theory for deep learning via over-parameterization","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Allen-Zhu"},{"key":"ref51","article-title":"Convergence of implicit gradient descent for training two-layer physics-informed neural networks","author":"Xu","year":"2024","journal-title":"arXiv:2407.02827"},{"key":"ref52","article-title":"Convergence analysis of natural gradient descent for over-parameterized physics-informed neural networks","author":"Xu","year":"2024","journal-title":"arXiv:2408.00573"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.1016\/j.neucom.2025.129414"},{"key":"ref54","doi-asserted-by":"publisher","DOI":"10.2139\/ssrn.4835325"},{"key":"ref55","article-title":"FC-KAN: Function combinations in Kolmogorov\u2013Arnold networks","author":"Ta","year":"2024","journal-title":"arXiv:2409.01763"},{"key":"ref56","article-title":"Exploring the potential of polynomial basis functions in Kolmogorov\u2013Arnold networks: A comparative study of different groups of polynomials","author":"Teymoor Seydi","year":"2024","journal-title":"arXiv:2406.02583"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2015.123"},{"key":"ref58","first-page":"1","article-title":"Self-normalizing neural networks","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"30","author":"Klambauer"},{"key":"ref59","first-page":"1139","article-title":"On the importance of initialization and momentum in deep learning","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Sutskever"},{"key":"ref60","article-title":"Accurate, large minibatch SGD: Training ImageNet in 1 hour","author":"Goyal","year":"2017","journal-title":"arXiv:1706.02677"},{"key":"ref61","first-page":"948","article-title":"Understanding gradient descent on the edge of stability in deep learning","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Arora"},{"key":"ref62","first-page":"1","article-title":"Minibatch vs local SGD with shuffling: Tight convergence bounds and beyond","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Yun"},{"key":"ref63","first-page":"1","article-title":"Don\u2019t use large mini-batches, use local SGD","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Lin"},{"key":"ref64","article-title":"On exponential convergence of SGD in non-convex over-parametrized learning","author":"Bassily","year":"2018","journal-title":"arXiv:1811.02564"},{"key":"ref65","first-page":"41852","article-title":"The challenges of the nonlinear regime for physics-informed neural networks","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"37","author":"Bonfanti"},{"key":"ref66","doi-asserted-by":"publisher","DOI":"10.1093\/imaiai\/iaac012"}],"container-title":["IEEE Transactions on Information Theory"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/18\/11134633\/11079726.pdf?arnumber=11079726","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,8,25]],"date-time":"2025-08-25T20:44:42Z","timestamp":1756154682000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11079726\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,9]]},"references-count":66,"journal-issue":{"issue":"9"},"URL":"https:\/\/doi.org\/10.1109\/tit.2025.3588401","relation":{},"ISSN":["0018-9448","1557-9654"],"issn-type":[{"value":"0018-9448","type":"print"},{"value":"1557-9654","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,9]]}}}