{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,23]],"date-time":"2026-10-23T20:22:16Z","timestamp":1792786936658,"version":"4.5.0"},"reference-count":85,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"11","license":[{"start":{"date-parts":[[2026,11,1]],"date-time":"2026-11-01T00:00:00Z","timestamp":1793491200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,11,1]],"date-time":"2026-11-01T00:00:00Z","timestamp":1793491200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,11,1]],"date-time":"2026-11-01T00:00:00Z","timestamp":1793491200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"name":"National Natural Science Foundation of China","award":["62372132"],"award-info":[{"award-number":["62372132"]}],"id":[{"id":"https:\/\/ror.org\/01h0zpd94","id-type":"ROR","asserted-by":"crossref"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Pattern Anal. Mach. Intell."],"published-print":{"date-parts":[[2026,11]]},"DOI":"10.1109\/tpami.2026.3723414","type":"journal-article","created":{"date-parts":[[2026,8,17]],"date-time":"2026-08-17T19:07:11Z","timestamp":1786993631000},"page":"15412-15426","source":"Crossref","is-referenced-by-count":0,"title":["PolaFormer++: Polarity-Aware Linear Attention With Channel-Wise Spikiness"],"prefix":"10.1109","volume":"48","author":[{"ORCID":"https:\/\/orcid.org\/0009-0003-5851-1727","authenticated-orcid":false,"given":"Weikang","family":"Meng","sequence":"first","affiliation":[{"name":"Harbin Institute of Technology","place":["Shenzhen, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6272-2971","authenticated-orcid":false,"given":"Yadan","family":"Luo","sequence":"additional","affiliation":[{"name":"The University of Queensland","place":["St Lucia, QLD, Australia"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Liangyu","family":"Huo","sequence":"additional","affiliation":[{"name":"Harbin Institute of Technology","place":["Shenzhen, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0653-4535","authenticated-orcid":false,"given":"Yingjian","family":"Li","sequence":"additional","affiliation":[{"name":"Pengcheng Laboratory","place":["Shenzhen, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2197-9038","authenticated-orcid":false,"given":"Yaowei","family":"Wang","sequence":"additional","affiliation":[{"name":"Harbin Institute of Technology","place":["Shenzhen, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1470-6998","authenticated-orcid":false,"given":"Zheng","family":"Zhang","sequence":"additional","affiliation":[{"name":"Harbin Institute of Technology","place":["Shenzhen, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2999-2088","authenticated-orcid":false,"given":"Heng Tao","family":"Shen","sequence":"additional","affiliation":[{"name":"Tongji University","place":["Shanghai, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","article-title":"An image is worth 16x16 words: Transformers for image recognition at scale","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Dosovitskiy","year":"2021"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.01587"},{"key":"ref3","first-page":"7281","article-title":"HRFormer: High-resolution vision transformer for dense predict","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Yuan","year":"2021"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00986"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00061"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1007\/s41095-022-0274-8"},{"key":"ref7","article-title":"Reformer: The efficient transformer","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Kitaev","year":"2020"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52734.2025.01218"},{"key":"ref9","first-page":"5156","article-title":"Transformers are RNNs: Fast autoregressive transformers with linear attention","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Katharopoulos","year":"2020"},{"key":"ref10","article-title":"CosFormer: Rethinking softmax in attention","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Qin","year":"2022"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v39i5.32550"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52734.2025.02353"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00548"},{"key":"ref14","first-page":"53551","article-title":"The hedgehog & the porcupine: Expressive linear attentions with softmax mimicry","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Zhang","year":"2024"},{"key":"ref15","first-page":"41518","article-title":"Linear log-normal attention with unbiased concentration","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Nahshan","year":"2024"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.01176"},{"key":"ref17","first-page":"50020","article-title":"PolaFormer: Polarity-aware linear attention for vision transformers","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Meng","year":"2025"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2009.5206848"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-10602-1_48"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.350"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-018-1140-0"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW.2017.150"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.5244\/C.26.135"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-27413-8_47"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2001.937655"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2015.7299156"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1007\/s11042-016-4020-z"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1145\/3306346.3322980"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.1706.03762"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.52202\/079017-3273"},{"key":"ref31","first-page":"62429","article-title":"Vision mamba: Efficient visual representation learning with bidirectional state space model","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Zhu","year":"2024"},{"key":"ref32","first-page":"83166","article-title":"Vision-RWKV: Efficient and scalable visual perception with RWKV-like architectures","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Duan","year":"2025"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1189"},{"key":"ref34","first-page":"2122","article-title":"SkyFormer: Remodel self-attention with Gaussian Kernel and Nystr\u00f6m method","volume-title":"Proc. Annu. Conf. Neural Inf. Process. Syst.","author":"Chen","year":"2021"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP39728.2021.9413493"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-024-02035-5"},{"key":"ref37","article-title":"Rethinking attention with performers","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Choromanski","year":"2021"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1109\/WACV48630.2021.00357"},{"key":"ref39","article-title":"Retentive network: A successor to transformer for large language models","author":"Sun","year":"2023"},{"key":"ref40","article-title":"Lightning attention-2: A free lunch for handling unlimited sequence lengths in large language models","author":"Qin","year":"2024"},{"key":"ref41","first-page":"24226","article-title":"FlowFormer: Linearizing transformers with conservation flows","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Wu","year":"2022"},{"key":"ref42","article-title":"Scaling transnormer to 175 billion parameters","author":"Qin","year":"2023"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P19-1146"},{"key":"ref44","first-page":"1614","article-title":"From softmax to sparsemax: A sparse model of attention and multi-label classification","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Martins","year":"2016"},{"key":"ref45","first-page":"144","article-title":"Differential transformer","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Ye","year":"2025"},{"key":"ref46","article-title":"Norm \u00d7 direction: Restoring the missing query norm in vision linear attention","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Meng","year":"2026"},{"key":"ref47","first-page":"56501","article-title":"Gated linear attention transformers with hardware-efficient training","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Yang","year":"2024"},{"key":"ref48","first-page":"597","article-title":"On the computational complexity of self-attention","volume-title":"Proc. Int. Conf. Algorithmic Learn. Theory","author":"Keles","year":"2023"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1007\/0-387-31077-0"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.00539"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01169"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01157"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.01598"},{"key":"ref54","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72973-7_8"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-91979-4_2"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52734.2025.00423"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2024.3401450"},{"key":"ref58","first-page":"37393","article-title":"MogaNet: Multi-order gated aggregation network","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Li","year":"2024"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.52202\/079017-4039"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.01792"},{"key":"ref61","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-20053-3_27"},{"key":"ref62","first-page":"12633","article-title":"Global context vision transformers","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Hatamizadeh","year":"2023"},{"key":"ref63","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.324"},{"key":"ref64","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.322"},{"key":"ref65","article-title":"MMCV: OpenMMLab computer vision foundation","author":"Contributors","year":"2018"},{"key":"ref66","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.00714"},{"key":"ref67","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01186"},{"key":"ref68","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01385"},{"key":"ref69","first-page":"20014","article-title":"XCiT: Cross-covariance image transformers","volume-title":"Proc. Annu. Conf. Neural Inf. Process. Syst.","author":"Ali","year":"2021"},{"key":"ref70","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.00599"},{"key":"ref71","doi-asserted-by":"publisher","DOI":"10.52202\/068431-0084"},{"key":"ref72","first-page":"12077","article-title":"SegFormer: Simple and efficient design for semantic segmentation with transformers","volume-title":"Proc. Annu. Conf. Neural Inf. Process. Syst.","author":"Xie","year":"2021"},{"key":"ref73","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00387"},{"key":"ref74","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52734.2025.00718"},{"key":"ref75","first-page":"12225","article-title":"MHLA: Restoring expressivity of linear attention via token-level multi-head","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Zhang","year":"2026"},{"key":"ref76","article-title":"Is attention all that NeRF needs","volume-title":"Proc. Int. Conf. Learn. Representations","author":"V. T.","year":"2023"},{"key":"ref77","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW56347.2022.00061"},{"key":"ref78","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.01174"},{"key":"ref79","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52734.2025.02619"},{"key":"ref80","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51701.2025.02268"},{"key":"ref81","doi-asserted-by":"publisher","DOI":"10.1109\/ICCVW54120.2021.00210"},{"key":"ref82","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72649-1_13"},{"key":"ref83","doi-asserted-by":"publisher","DOI":"10.52202\/079017-2515"},{"key":"ref84","doi-asserted-by":"publisher","DOI":"10.52202\/085713-2046"},{"key":"ref85","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v40i26.39303"}],"container-title":["IEEE Transactions on Pattern Analysis and Machine Intelligence"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/34\/11723300\/11657470.pdf?arnumber=11657470","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,8]],"date-time":"2026-10-08T19:09:03Z","timestamp":1791486543000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11657470\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,11]]},"references-count":85,"journal-issue":{"issue":"11"},"URL":"https:\/\/doi.org\/10.1109\/tpami.2026.3723414","relation":{},"ISSN":["0162-8828","2160-9292","1939-3539"],"issn-type":[{"value":"0162-8828","type":"print"},{"value":"2160-9292","type":"electronic"},{"value":"1939-3539","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,11]]}}}