{"id":"https://openalex.org/W7166869770","doi":"https://doi.org/10.18653/v1/2026.findings-acl.348","title":"Rectified Sparse Attention for Efficient Long-Sequence Generation","display_name":"Rectified Sparse Attention for Efficient Long-Sequence Generation","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166869770","doi":"https://doi.org/10.18653/v1/2026.findings-acl.348"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.findings-acl.348","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.348","pdf_url":"https://aclanthology.org/2026.findings-acl.348.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.findings-acl.348.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139769468","display_name":"Yutao Sun","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yutao Sun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139747561","display_name":"Tianzhu Ye","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tianzhu Ye","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139790469","display_name":"Li Dong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li Dong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139727334","display_name":"Yuqing Xia","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yuqing Xia","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139762147","display_name":"Jian Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jian Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139796878","display_name":"Yizhao Gao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yizhao Gao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139828203","display_name":"Shijie Cao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shijie Cao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139724406","display_name":"Jianyong Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jianyong Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139803669","display_name":"Furu Wei","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Furu Wei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.87943848,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"7023","last_page":"7034"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.12639999389648438,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.12639999389648438,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.07050000131130219,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10860","display_name":"Speech and Audio Processing","score":0.05380000174045563,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/noise","display_name":"Noise (video)","score":0.29649999737739563},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.2721000015735626},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.2538999915122986},{"id":"https://openalex.org/keywords/focus","display_name":"Focus (optics)","score":0.2298000007867813},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.22939999401569366}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5931000113487244},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4691999852657318},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.29649999737739563},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.2721000015735626},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.2540999948978424},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.2538999915122986},{"id":"https://openalex.org/C192209626","wikidata":"https://www.wikidata.org/wiki/Q190909","display_name":"Focus (optics)","level":2,"score":0.2298000007867813},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.22939999401569366},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.2272000014781952},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.22139999270439148}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.findings-acl.348","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.348","pdf_url":"https://aclanthology.org/2026.findings-acl.348.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.findings-acl.348","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.348","pdf_url":"https://aclanthology.org/2026.findings-acl.348.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G5088477901","display_name":null,"funder_award_id":"62272264","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G6777344962","display_name":null,"funder_award_id":"2020YFA0804503","funder_id":"https://openalex.org/F4320335777","funder_display_name":"National Key Research and Development Program of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320335777","display_name":"National Key Research and Development Program of China","ror":null}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166869770.pdf","grobid_xml":"https://content.openalex.org/works/W7166869770.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Efficient":[0],"long-sequence":[1],"generation":[2,29,88],"is":[3],"a":[4,39,60,109],"critical":[5],"challenge":[6],"for":[7,112],"Large":[8],"Language":[9],"Models.While":[10],"recent":[11],"sparse":[12],"decoding":[13,102],"methods":[14],"improve":[15],"efficiency,":[16],"they":[17],"suffer":[18],"from":[19],"KV":[20,54],"cache":[21,55],"misalignment,":[22],"where":[23],"approximation":[24],"errors":[25],"accumulate":[26],"and":[27,68,80],"degrade":[28],"quality.In":[30],"this":[31],"work,":[32],"we":[33],"propose":[34],"Rectified":[35],"Sparse":[36],"Attention":[37],"(ReSA),":[38],"simple":[40],"yet":[41],"effective":[42],"method":[43],"that":[44,84],"combines":[45],"block-sparse":[46],"attention":[47],"with":[48,71,90],"periodic":[49],"dense":[50,61],"rectification.By":[51],"refreshing":[52],"the":[53,72],"at":[56,103],"fixed":[57],"intervals":[58],"using":[59],"forward":[62],"pass,":[63],"ReSA":[64,85,94],"bounds":[65],"error":[66],"accumulation":[67],"preserves":[69],"alignment":[70],"pretraining":[73],"distribution.Experiments":[74],"across":[75],"math":[76],"reasoning,":[77],"language":[78],"modeling,":[79],"retrieval":[81],"tasks":[82],"demonstrate":[83],"achieves":[86],"near-lossless":[87],"quality":[89],"significantly":[91],"improved":[92],"efficiency.Notably,":[93],"delivers":[95],"up":[96],"to":[97],"3.77\u00d7":[98],"end-to-end":[99],"speedup":[100],"under":[101],"256K":[104],"sequence":[105],"length,":[106],"making":[107],"it":[108],"practical":[110],"solution":[111],"scalable":[113],"long-context":[114],"inference.":[115]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-02T00:00:00"}