{"id":"https://openalex.org/W7164145500","doi":"https://doi.org/10.48550/arxiv.2606.10896","title":"Flash-GMM: A Memory-Efficient Kernel for Scalable Soft Clustering","display_name":"Flash-GMM: A Memory-Efficient Kernel for Scalable Soft Clustering","publication_year":2026,"publication_date":"2026-06-09","ids":{"openalex":"https://openalex.org/W7164145500","doi":"https://doi.org/10.48550/arxiv.2606.10896"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.10896","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.10896","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.10896","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5120024082","display_name":"Gal Bloch","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bloch, Gal","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138308442","display_name":"Ariel Gera","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gera, Ariel","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5078223247","display_name":"Matan Orbach","orcid":"https://orcid.org/0000-0001-9923-5093"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Orbach, Matan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138326091","display_name":"Ohad Eytan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Eytan, Ohad","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5072770537","display_name":"Assaf Toledo","orcid":"https://orcid.org/0000-0002-5309-267X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Toledo, Assaf","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":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11901","display_name":"Bayesian Methods and Mixture Models","score":0.599399983882904,"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"}},"topics":[{"id":"https://openalex.org/T11901","display_name":"Bayesian Methods and Mixture Models","score":0.599399983882904,"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/T10637","display_name":"Advanced Clustering Algorithms Research","score":0.06610000133514404,"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/T10057","display_name":"Face and Expression Recognition","score":0.031599998474121094,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/kernel","display_name":"Kernel (algebra)","score":0.6906999945640564},{"id":"https://openalex.org/keywords/cluster-analysis","display_name":"Cluster analysis","score":0.6825000047683716},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.6294999718666077},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.6039999723434448},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.5925999879837036},{"id":"https://openalex.org/keywords/mixture-model","display_name":"Mixture model","score":0.592199981212616},{"id":"https://openalex.org/keywords/gaussian-process","display_name":"Gaussian process","score":0.3522999882698059},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.3395000100135803},{"id":"https://openalex.org/keywords/implementation","display_name":"Implementation","score":0.32710000872612}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7289000153541565},{"id":"https://openalex.org/C74193536","wikidata":"https://www.wikidata.org/wiki/Q574844","display_name":"Kernel (algebra)","level":2,"score":0.6906999945640564},{"id":"https://openalex.org/C73555534","wikidata":"https://www.wikidata.org/wiki/Q622825","display_name":"Cluster analysis","level":2,"score":0.6825000047683716},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.6294999718666077},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.6039999723434448},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.5925999879837036},{"id":"https://openalex.org/C61224824","wikidata":"https://www.wikidata.org/wiki/Q2260434","display_name":"Mixture model","level":2,"score":0.592199981212616},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.428600013256073},{"id":"https://openalex.org/C61326573","wikidata":"https://www.wikidata.org/wiki/Q1496376","display_name":"Gaussian process","level":3,"score":0.3522999882698059},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3395000100135803},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3384999930858612},{"id":"https://openalex.org/C26713055","wikidata":"https://www.wikidata.org/wiki/Q245962","display_name":"Implementation","level":2,"score":0.32710000872612},{"id":"https://openalex.org/C122280245","wikidata":"https://www.wikidata.org/wiki/Q620622","display_name":"Kernel method","level":3,"score":0.32499998807907104},{"id":"https://openalex.org/C106487976","wikidata":"https://www.wikidata.org/wiki/Q685816","display_name":"Matrix (chemical analysis)","level":2,"score":0.31790000200271606},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.2964000105857849},{"id":"https://openalex.org/C81669768","wikidata":"https://www.wikidata.org/wiki/Q2359161","display_name":"Precision and recall","level":2,"score":0.29030001163482666},{"id":"https://openalex.org/C163716315","wikidata":"https://www.wikidata.org/wiki/Q901177","display_name":"Gaussian","level":2,"score":0.28220000863075256},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.27720001339912415},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2741999924182892},{"id":"https://openalex.org/C2777210771","wikidata":"https://www.wikidata.org/wiki/Q4927124","display_name":"Block (permutation group theory)","level":2,"score":0.27239999175071716},{"id":"https://openalex.org/C105611402","wikidata":"https://www.wikidata.org/wiki/Q2976589","display_name":"Spectral clustering","level":3,"score":0.2671000063419342},{"id":"https://openalex.org/C179799912","wikidata":"https://www.wikidata.org/wiki/Q205084","display_name":"Computational complexity theory","level":2,"score":0.2623000144958496},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.2581000030040741},{"id":"https://openalex.org/C182335926","wikidata":"https://www.wikidata.org/wiki/Q17093020","display_name":"Kernel principal component analysis","level":4,"score":0.25589999556541443},{"id":"https://openalex.org/C207968372","wikidata":"https://www.wikidata.org/wiki/Q310401","display_name":"k-means clustering","level":3,"score":0.2556999921798706},{"id":"https://openalex.org/C7218915","wikidata":"https://www.wikidata.org/wiki/Q1054475","display_name":"Gaussian function","level":3,"score":0.2538999915122986},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.25029999017715454}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.10896","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.10896","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.10896","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.10896","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"We":[0,76,126],"present":[1],"\\textbf{Flash-GMM},":[2],"a":[3,19,38,84],"fused":[4],"Triton":[5],"kernel":[6,129],"for":[7,71,88],"efficient":[8],"computation":[9],"of":[10],"Gaussian":[11],"Mixture":[12],"Models":[13],"(GMMs)":[14],"over":[15,41],"large-scale":[16],"data":[17],"in":[18,33],"single":[20],"GPU":[21,34],"pass.":[22],"By":[23],"eliminating":[24],"the":[25,29,67,128],"need":[26],"to":[27,97,101,112],"materialize":[28],"full":[30],"responsibility":[31],"matrix":[32],"memory,":[35],"Flash-GMM":[36,65],"achieves":[37],"\\textbf{20$\\times$}":[39],"speedup":[40],"existing":[42],"implementations":[43],"and":[44,90],"enables":[45],"training":[46],"on":[47,56],"datasets":[48],"more":[49],"than":[50,53],"\\textbf{100$\\times$}":[51],"larger":[52],"previously":[54],"feasible":[55],"one":[57],"device.":[58],"To":[59],"demonstrate":[60],"its":[61],"impact,":[62],"we":[63],"integrate":[64],"into":[66],"IVF":[68],"coarse":[69],"quantizer":[70],"approximate":[72],"nearest-neighbor":[73],"(ANN)":[74],"search.":[75],"show":[77],"that":[78,91],"soft":[79],"GMM":[80,92],"clustering":[81],"is":[82],"now":[83],"viable":[85],"drop-in":[86],"replacement":[87],"$k$-means,":[89],"responsibilities":[93],"can":[94],"be":[95],"leveraged":[96],"assign":[98],"border":[99],"vectors":[100],"multiple":[102],"clusters.":[103],"Our":[104],"approach":[105],"reaches":[106],"fixed":[107],"recall":[108],"targets":[109],"with":[110],"up":[111],"$1.7\\times$":[113],"fewer":[114],"distance":[115],"computations,":[116],"or":[117],"equivalently,":[118],"yields":[119],"$+2$--$12$":[120],"recall@10":[121],"at":[122],"matched":[123],"computational":[124],"cost.":[125],"release":[127],"as":[130],"an":[131],"open-source":[132],"project.":[133]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-06-11T00:00:00"}