{"id":"https://openalex.org/W7166808789","doi":"https://doi.org/10.18653/v1/2026.acl-long.1732","title":"Resonating with RoPE: Spectral Quantization for High-Fidelity Key Cache Compression","display_name":"Resonating with RoPE: Spectral Quantization for High-Fidelity Key Cache Compression","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166808789","doi":"https://doi.org/10.18653/v1/2026.acl-long.1732"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2026.acl-long.1732","doi":"https://doi.org/10.18653/v1/2026.acl-long.1732","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.1732","pdf_url":"https://aclanthology.org/2026.acl-long.1732.pdf","source":{"id":"https://openalex.org/S4393915740","display_name":"Proceedings of the conference - Association for Computational Linguistics. Meeting","issn_l":"0736-587X","issn":["0736-587X"],"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":["norway-1"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://aclanthology.org/2026.acl-long.1732.pdf","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139773339","display_name":"Xuefei Wang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I154099455","display_name":"Shandong University","ror":"https://ror.org/0207yh398","country_code":"CN","type":"education","lineage":["https://openalex.org/I154099455"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xuefei Wang","raw_affiliation_strings":["Shandong University ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shandong University ,","institution_ids":["https://openalex.org/I154099455"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101669572","display_name":"Haoyu Tang","orcid":"https://orcid.org/0000-0001-7095-9688","observed_orcids":["https://orcid.org/0000-0001-7095-9688","https://orcid.org/0009-0003-9634-4034"]},"institutions":[{"id":"https://openalex.org/I154099455","display_name":"Shandong University","ror":"https://ror.org/0207yh398","country_code":"CN","type":"education","lineage":["https://openalex.org/I154099455"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Haoyu Tang","raw_affiliation_strings":["Shandong University ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shandong University ,","institution_ids":["https://openalex.org/I154099455"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129703518","display_name":"Tianyuan Liang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I154099455","display_name":"Shandong University","ror":"https://ror.org/0207yh398","country_code":"CN","type":"education","lineage":["https://openalex.org/I154099455"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Tianyuan Liang","raw_affiliation_strings":["Shandong University ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shandong University ,","institution_ids":["https://openalex.org/I154099455"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139766463","display_name":"Zhibin Wang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I154099455","display_name":"Shandong University","ror":"https://ror.org/0207yh398","country_code":"CN","type":"education","lineage":["https://openalex.org/I154099455"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhibin Wang","raw_affiliation_strings":["Shandong University ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shandong University ,","institution_ids":["https://openalex.org/I154099455"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5069741536","display_name":"Yupeng Hu","orcid":"https://orcid.org/0000-0002-7358-7426","observed_orcids":["https://orcid.org/0000-0002-7358-7426"]},"institutions":[{"id":"https://openalex.org/I154099455","display_name":"Shandong University","ror":"https://ror.org/0207yh398","country_code":"CN","type":"education","lineage":["https://openalex.org/I154099455"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yupeng Hu","raw_affiliation_strings":["Shandong University ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shandong University ,","institution_ids":["https://openalex.org/I154099455"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5139737086","display_name":"Weili Guan","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Weili Guan","raw_affiliation_strings":["Harbin Institute of Technology (Shenzhen)"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Harbin Institute of Technology (Shenzhen)","institution_ids":["https://openalex.org/I204983213"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"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":0.47183555,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"37328","last_page":"37348"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.8047000169754028,"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":0.8047000169754028,"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.06700000166893005,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/T10028","display_name":"Topic Modeling","score":0.016899999231100082,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/kv-cache-compression","display_name":"KV cache compression","score":0.800000011920929},{"id":"https://openalex.org/keywords/key-value-cache","display_name":"key-value cache","score":0.5830000042915344},{"id":"https://openalex.org/keywords/rope","display_name":"rope","score":0.6240000128746033},{"id":"https://openalex.org/keywords/spectral-quantization","display_name":"spectral quantization","score":0.6159999966621399},{"id":"https://openalex.org/keywords/discrete-cosine-transform","display_name":"discrete cosine transform","score":0.890999972820282},{"id":"https://openalex.org/keywords/frequency-domain-representation","display_name":"frequency-domain representation","score":0.765999972820282},{"id":"https://openalex.org/keywords/long-context-llms","display_name":"long-context LLMs","score":0.7459999918937683},{"id":"https://openalex.org/keywords/triton-kernels","display_name":"Triton kernels","score":0.656000018119812},{"id":"https://openalex.org/keywords/memory-footprint-reduction","display_name":"memory footprint reduction","score":0.6430000066757202}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5940999984741211},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.48030000925064087},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.448199987411499},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.4341000020503998},{"id":"https://openalex.org/C78548338","wikidata":"https://www.wikidata.org/wiki/Q2493","display_name":"Data compression","level":2,"score":0.4196000099182129},{"id":"https://openalex.org/C180016635","wikidata":"https://www.wikidata.org/wiki/Q2712821","display_name":"Compression (physics)","level":2,"score":0.3937000036239624},{"id":"https://openalex.org/C199833920","wikidata":"https://www.wikidata.org/wiki/Q612536","display_name":"Vector quantization","level":2,"score":0.334199994802475},{"id":"https://openalex.org/C104267543","wikidata":"https://www.wikidata.org/wiki/Q208163","display_name":"Signal processing","level":3,"score":0.28690001368522644},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.2775000035762787},{"id":"https://openalex.org/C81081738","wikidata":"https://www.wikidata.org/wiki/Q55542","display_name":"Lossless compression","level":3,"score":0.2574999928474426},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.2567000091075897},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.25270000100135803},{"id":"https://openalex.org/C24326235","wikidata":"https://www.wikidata.org/wiki/Q126095","display_name":"Electronic engineering","level":1,"score":0.25119999051094055}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.acl-long.1732","doi":"https://doi.org/10.18653/v1/2026.acl-long.1732","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.1732","pdf_url":"https://aclanthology.org/2026.acl-long.1732.pdf","source":{"id":"https://openalex.org/S4393915740","display_name":"Proceedings of the conference - Association for Computational Linguistics. Meeting","issn_l":"0736-587X","issn":["0736-587X"],"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":["norway-1"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.acl-long.1732","doi":"https://doi.org/10.18653/v1/2026.acl-long.1732","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.1732","pdf_url":"https://aclanthology.org/2026.acl-long.1732.pdf","source":{"id":"https://openalex.org/S4393915740","display_name":"Proceedings of the conference - Association for Computational Linguistics. Meeting","issn_l":"0736-587X","issn":["0736-587X"],"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":["norway-1"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166808789.pdf","grobid_xml":"https://content.openalex.org/works/W7166808789.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7202166650","https://openalex.org/W7165750449","https://openalex.org/W7167702318","https://openalex.org/W4402671912","https://openalex.org/W7102430828","https://openalex.org/W7161811629","https://openalex.org/W4396912812","https://openalex.org/W4415085075","https://openalex.org/W7172426198","https://openalex.org/W6929418818"],"abstract_inverted_index":{"The":[0],"linear":[1],"growth":[2],"of":[3],"KV":[4],"cache":[5,14],"bottlenecks":[6],"long-context":[7],"LLMs,":[8],"yet":[9],"RoPE-induced":[10],"oscillations":[11,35],"complicate":[12],"Key":[13],"quantization.To":[15],"address":[16],"this":[17],"issue,":[18],"we":[19,62],"propose":[20],"SpectrumQuant,":[21],"a":[22],"frequency-domain":[23],"framework":[24],"that":[25],"utilizes":[26],"the":[27],"Discrete":[28],"Cosine":[29],"Transform":[30],"(DCT)":[31],"to":[32,52,89],"convert":[33],"these":[34],"into":[36],"sparse":[37,73],"spectral":[38],"representations.Specifically,":[39],"our":[40],"pipeline":[41],"integrates":[42],"dominant":[43],"frequency":[44],"extraction,":[45],"hybrid":[46],"bit-width":[47],"allocation,":[48],"and":[49,71,86],"high-frequency":[50],"preemphasis":[51],"maximize":[53],"fidelity":[54],"while":[55],"minimizing":[56],"memory":[57],"footprint.To":[58],"eliminate":[59],"computational":[60],"overhead,":[61],"develop":[63],"fused":[64],"Triton":[65],"kernels":[66],"featuring":[67],"deferred":[68],"inverse":[69],"transformation":[70],"on-chip":[72],"accumulation.Extensive":[74],"experiments":[75],"on":[76],"several":[77],"benchmarks":[78],"confirm":[79],"SpectrumQuant":[80],"achieves":[81],"efficient":[82],"compression":[83],"with":[84],"performance":[85],"latency":[87],"comparable":[88],"FP16":[90],"baselines.":[91]},"counts_by_year":[],"updated_date":"2026-10-29T09:56:32.521200","created_date":"2026-07-02T00:00:00"}
