{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T05:44:37Z","timestamp":1782798277478,"version":"3.54.5"},"reference-count":39,"publisher":"IEEE","license":[{"start":{"date-parts":[[2026,5,18]],"date-time":"2026-05-18T00:00:00Z","timestamp":1779062400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,5,18]],"date-time":"2026-05-18T00:00:00Z","timestamp":1779062400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,5,18]]},"DOI":"10.1109\/infocom59046.2026.11571570","type":"proceedings-article","created":{"date-parts":[[2026,6,29]],"date-time":"2026-06-29T19:38:15Z","timestamp":1782761895000},"page":"1-10","source":"Crossref","is-referenced-by-count":0,"title":["HierMoE: Accelerating MoE Training with Hierarchical Token Deduplication and Expert Swap"],"prefix":"10.1109","author":[{"given":"Wenxiang","family":"Lin","sequence":"first","affiliation":[{"name":"Harbin Institute of Technology,School of Computer Science and Technology,Shenzhen"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xinglin","family":"Pan","sequence":"additional","affiliation":[{"name":"The Hong Kong University of Science and Technology (Guangzhou),Data Science and Analytics Thrust"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lin","family":"Zhang","sequence":"additional","affiliation":[{"name":"The Hong Kong University of Science and Technology,Department of Computer Science and Engineering"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shaohuai","family":"Shi","sequence":"additional","affiliation":[{"name":"Harbin Institute of Technology,School of Computer Science and Technology,Shenzhen"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xuan","family":"Wang","sequence":"additional","affiliation":[{"name":"Harbin Institute of Technology,School of Computer Science and Technology,Shenzhen"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiaowen","family":"Chu","sequence":"additional","affiliation":[{"name":"The Hong Kong University of Science and Technology (Guangzhou),Data Science and Analytics Thrust"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Outrageously large neural networks: The sparsely-gated mixture-of-experts layer","volume-title":"International Conference on Learning Representations","author":"Shazeer"},{"key":"ref2","article-title":"Gshard: Scaling giant models with conditional computation and automatic sharding","volume-title":"International Conference on Learning Representations","author":"Lepikhin"},{"key":"ref3","article-title":"Mixtral of experts","author":"Jiang","year":"2024"},{"key":"ref4","article-title":"Qwen3 technical report","author":"Yang","year":"2025"},{"key":"ref5","article-title":"Deepseek-v2: A strong, economical, and efficient mixture-of-experts language model","year":"2024"},{"key":"ref6","article-title":"Deepseek-v3 technical report","author":"Liu","year":"2024"},{"key":"ref7","article-title":"Tutel: Adaptive mixture-of-experts at scale","volume-title":"Proceedings of Machine Learning and Systems","volume":"5","author":"Hwang"},{"key":"ref8","first-page":"13 782","article-title":"Gating dropout: Communication-efficient regularization for sparsely activated transformers","volume-title":"International Conference on Machine Learning","author":"Liu"},{"key":"ref9","first-page":"945","article-title":"Accelerating distributed MoE training and inference with lina","volume-title":"USENIX Annual Technical Conference","author":"Li"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1145\/3627703.3650083"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1145\/3669940.3707272"},{"key":"ref12","first-page":"6265","article-title":"BASE layers: Simplifying training of large, sparse models","volume-title":"International Conference on Machine Learning","author":"Lewis"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.52202\/068431-0515"},{"issue":"1","key":"ref14","first-page":"5232","article-title":"Switch transformers: Scaling to trillion parameter models with simple and efficient sparsity","volume":"23","author":"Fedus","year":"2022","journal-title":"The Journal of Machine Learning Research"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.52202\/068431-2507"},{"key":"ref16","article-title":"From sparse to soft mixtures of experts","author":"Puigcerver","year":"2023"},{"key":"ref17","first-page":"18 332","article-title":"Deepspeed-moe: Advancing mixture-of-experts inference and training to power next-generation ai scale","volume-title":"International Conference on Machine Learning. PMLR","author":"Rajbhandari"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/SC41404.2022.00051"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1145\/3503221.3508417"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1145\/3503221.3508418"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/INFOCOM53939.2023.10228874"},{"key":"ref22","first-page":"961","article-title":"SmartMoE: Efficiently training Sparsely-Activated models through combining offline and online parallelization","volume-title":"USENIX Annual Technical Conference","author":"Zhai"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1145\/3603269.3604869"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1109\/INFOCOM52122.2024.10621327"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1145\/3620666.3651379"},{"key":"ref26","first-page":"74","article-title":"Lancet: Accelerating mixture-of-experts training by overlapping weight gradient computation and all-to-all communication","volume-title":"Proceedings of Machine Learning and Systems","volume":"6","author":"Jiang"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1109\/INFOCOM55648.2025.11044508"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1109\/ICDCS63083.2025.00063"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1109\/ICDCS63083.2025.00061"},{"key":"ref30","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-031-99872-0_23","article-title":"ScheInfer: Efficient inference of large language models with task scheduling on moderate gpus","volume-title":"European Conference on Parallel Processing","author":"Lin"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1145\/1851182.1851192"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1145\/1402946.1402967"},{"key":"ref33","first-page":"739","article-title":"Topoopt: Co-optimizing network topology and parallelization strategy for distributed training jobs","volume-title":"20th USENIX Symposium on Networked Systems Design and Implementation","author":"Wang"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1145\/3695053.3731412"},{"key":"ref35","article-title":"Large scale distributed deep networks","volume":"25","author":"Dean","year":"2012","journal-title":"Advances in neural information processing systems"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1109\/MNET.011.2000530"},{"key":"ref37","first-page":"401","article-title":"Towards scalable distributed training of deep learning on public cloud clusters","volume-title":"Proceedings of Machine Learning and Systems","volume":"3","author":"Shi"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1145\/3588964"},{"key":"ref39","first-page":"1501","article-title":"l1, p-norm regularization: Error bounds and convergence rate analysis of first-order methods","volume-title":"International conference on machine learning","author":"Zhou"}],"event":{"name":"IEEE INFOCOM 2026 - IEEE Conference on Computer Communications","location":"Tokyo, Japan","start":{"date-parts":[[2026,5,18]]},"end":{"date-parts":[[2026,5,21]]}},"container-title":["IEEE INFOCOM 2026 - IEEE Conference on Computer Communications"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11571071\/11571169\/11571570.pdf?arnumber=11571570","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T05:12:58Z","timestamp":1782796378000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11571570\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,18]]},"references-count":39,"URL":"https:\/\/doi.org\/10.1109\/infocom59046.2026.11571570","relation":{},"subject":[],"published":{"date-parts":[[2026,5,18]]}}}