{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,29]],"date-time":"2026-05-29T09:53:28Z","timestamp":1780048408959,"version":"3.53.1"},"reference-count":38,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","license":[{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"am","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"name":"NSF","award":["ECCS-2409701"],"award-info":[{"award-number":["ECCS-2409701"]}]},{"name":"NSF","award":["IIS-2402952"],"award-info":[{"award-number":["IIS-2402952"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Signal Process. Lett."],"published-print":{"date-parts":[[2026]]},"DOI":"10.1109\/lsp.2026.3684818","type":"journal-article","created":{"date-parts":[[2026,4,16]],"date-time":"2026-04-16T19:56:26Z","timestamp":1776369386000},"page":"1661-1665","source":"Crossref","is-referenced-by-count":2,"title":["On the Convergence of Gradient Descent on Learning Transformers With Residual Connections"],"prefix":"10.1109","volume":"33","author":[{"ORCID":"https:\/\/orcid.org\/0009-0000-0388-8840","authenticated-orcid":false,"given":"Zhen","family":"Qin","sequence":"first","affiliation":[{"name":"Michigan Institute for Computational Discovery and Engineering, Department of Electrical Engineering and Computer Science, Department of Statistics, University of Michigan, Ann Arbor, MI, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jinxin","family":"Zhou","sequence":"additional","affiliation":[{"name":"Department of Computer Science and Engineering, the Ohio State University, Columbus, OH, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jiachen","family":"Jiang","sequence":"additional","affiliation":[{"name":"Department of Computer Science and Engineering, the Ohio State University, Columbus, OH, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3856-0375","authenticated-orcid":false,"given":"Zhihui","family":"Zhu","sequence":"additional","affiliation":[{"name":"Department of Computer Science and Engineering, the Ohio State University, Columbus, OH, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.1706.03762"},{"issue":"8","key":"ref2","article-title":"Language models are unsupervised multitask learners","volume":"1","author":"Radford","year":"2019","journal-title":"OpenAI blog"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1145\/3219819.3219823"},{"key":"ref4","first-page":"15084","article-title":"Decision Transformer: Reinforcement learning via sequence modeling","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Chen","year":"2021"},{"key":"ref5","first-page":"1","article-title":"An image is worth 16 x 16 words: Transformers for image recognition at scale","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Dosovitskiy","year":"2020"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P19-1656"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/TCYB.2025.3556466"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/TVT.2023.3287530"},{"key":"ref9","article-title":"GPT-4 Technical Report","author":"Achiam","year":"2023"},{"key":"ref10","first-page":"86243","article-title":"Local to global: Learning dynamics and effect of initialization for transformers","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"37","author":"Makkuva","year":"2024"},{"key":"ref11","first-page":"20955","article-title":"From self-attention to Markov models: Unveiling the dynamics of generative transformers","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Ildiz","year":"2024"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.52202\/079017-1130"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.52202\/075280-2096"},{"key":"ref14","article-title":"A convergence analysis of gradient descent for deep linear neural networks","author":"Arora","year":"2018"},{"key":"ref15","first-page":"11961","article-title":"Global convergence of deep networks with one wide layer followed by pyramidal topology","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"2020","author":"Nguyen"},{"key":"ref16","article-title":"Convergence of gradient descent for deep neural networks","author":"Chatterjee","year":"2022"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/LSP.2024.3374085"},{"issue":"49","key":"ref18","first-page":"1","article-title":"Trained transformers learn linear models in-context","volume":"25","author":"Zhang","year":"2024","journal-title":"J. Mach. Learn. Res."},{"key":"ref19","article-title":"In-context convergence of transformers","author":"Huang","year":"2023"},{"key":"ref20","article-title":"In-context learning with representations: Contextual generalization of trained transformers","author":"Yang","year":"2024"},{"key":"ref21","article-title":"Training dynamics of in-context learning in linear attention","author":"Zhang","year":"2025"},{"key":"ref22","article-title":"Training nonlinear transformers for efficient in-context learning: A theoretical learning and generalization analysis","author":"Li","year":"2024"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.52202\/079017-2931"},{"key":"ref24","first-page":"52197","article-title":"On the convergence of encoder-only shallow transformers","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"36","author":"Wu","year":"2023"},{"key":"ref25","article-title":"Identity matters in deep learning","author":"Hardt","year":"2016"},{"key":"ref26","first-page":"7890","article-title":"Towards understanding the importance of shortcut connections in residual networks","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"32","author":"Liu","year":"2019"},{"key":"ref27","article-title":"Residual connections and normalization can provably prevent oversmoothing in GNNs","author":"Scholkemper","year":"2024"},{"key":"ref28","first-page":"27198","article-title":"Signal propagation in transformers: Theoretical perspectives and the role of rank collapse","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"35","author":"Noci","year":"2022"},{"key":"ref29","article-title":"Llama 2: Open foundation and fine-tuned chat models","author":"Touvron","year":"2023"},{"issue":"240","key":"ref30","first-page":"1","article-title":"PaLM: Scaling language modeling with pathways","volume":"24","author":"Chowdhery","year":"2023","journal-title":"J. Mach. Learn. Res."},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.acl-long.841"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/lsp.2026.3684818"},{"key":"ref33","article-title":"Timeseries forecasting for weather prediction","author":"Attri","year":"2020"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1002\/aisy.202300009"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1145\/3641142.3641143"},{"key":"ref36","article-title":"Residual connections harm generative representation learning","author":"Zhang","year":"2024"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D13-1170"},{"key":"ref38","first-page":"1","article-title":"Learning to adapt: In-context learning beyond stationarity","volume-title":"Proc. 14th Int. Conf. Learn. Representations","author":"Qin","year":"2026"}],"container-title":["IEEE Signal Processing Letters"],"original-title":[],"link":[{"URL":"https:\/\/ieeexplore.ieee.org\/ielam\/97\/11304147\/11482474-aam.pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/97\/11304147\/11482474.pdf?arnumber=11482474","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,4]],"date-time":"2026-05-04T19:39:24Z","timestamp":1777923564000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11482474\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026]]},"references-count":38,"URL":"https:\/\/doi.org\/10.1109\/lsp.2026.3684818","relation":{},"ISSN":["1070-9908","1558-2361"],"issn-type":[{"value":"1070-9908","type":"print"},{"value":"1558-2361","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026]]}}}