{"id":"https://openalex.org/W7155563963","doi":"https://doi.org/10.1145/3799830.3799837","title":"Hierarchical Sparse Plus Low Rank Compression of LLM","display_name":"Hierarchical Sparse Plus Low Rank Compression of LLM","publication_year":2025,"publication_date":"2025-12-17","ids":{"openalex":"https://openalex.org/W7155563963","doi":"https://doi.org/10.1145/3799830.3799837"},"language":null,"primary_location":{"id":"doi:10.1145/3799830.3799837","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3799830.3799837","pdf_url":null,"source":null,"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 13th ACM IKDD International Conference on Data Science","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3799830.3799837","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5067384528","display_name":"Pawan Kumar","orcid":"https://orcid.org/0000-0001-5632-6964"},"institutions":[{"id":"https://openalex.org/I64189192","display_name":"International Institute of Information Technology, Hyderabad","ror":"https://ror.org/05f11g639","country_code":"IN","type":"education","lineage":["https://openalex.org/I64189192"]}],"countries":["IN"],"is_corresponding":true,"raw_author_name":"pawan kumar","raw_affiliation_strings":["International Institute of Information Technology, Hyderabad, Hyderabad, India"],"raw_orcid":"https://orcid.org/0000-0001-5632-6964","affiliations":[{"raw_affiliation_string":"International Institute of Information Technology, Hyderabad, Hyderabad, India","institution_ids":["https://openalex.org/I64189192"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5081505598","display_name":"Aditi Gupta","orcid":"https://orcid.org/0000-0002-4496-9775"},"institutions":[{"id":"https://openalex.org/I64189192","display_name":"International Institute of Information Technology, Hyderabad","ror":"https://ror.org/05f11g639","country_code":"IN","type":"education","lineage":["https://openalex.org/I64189192"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Aditi Gupta","raw_affiliation_strings":["International Institute of Information Technology, Hyderabad, Hyderabad, India"],"raw_orcid":"https://orcid.org/0009-0009-3725-9482","affiliations":[{"raw_affiliation_string":"International Institute of Information Technology, Hyderabad, Hyderabad, India","institution_ids":["https://openalex.org/I64189192"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5067384528"],"corresponding_institution_ids":["https://openalex.org/I64189192"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.86607029,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"62","last_page":"70"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.19040000438690186,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.19040000438690186,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.11410000175237656,"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/T14347","display_name":"Big Data and Digital Economy","score":0.08869999647140503,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/compression","display_name":"Compression (physics)","score":0.5990999937057495},{"id":"https://openalex.org/keywords/sparse-matrix","display_name":"Sparse matrix","score":0.5795000195503235},{"id":"https://openalex.org/keywords/separable-space","display_name":"Separable space","score":0.5533999800682068},{"id":"https://openalex.org/keywords/permutation","display_name":"Permutation (music)","score":0.550000011920929},{"id":"https://openalex.org/keywords/factorization","display_name":"Factorization","score":0.516700029373169},{"id":"https://openalex.org/keywords/residual","display_name":"Residual","score":0.4964999854564667},{"id":"https://openalex.org/keywords/rank","display_name":"Rank (graph theory)","score":0.4959999918937683},{"id":"https://openalex.org/keywords/sequence","display_name":"Sequence (biology)","score":0.4909000098705292}],"concepts":[{"id":"https://openalex.org/C180016635","wikidata":"https://www.wikidata.org/wiki/Q2712821","display_name":"Compression (physics)","level":2,"score":0.5990999937057495},{"id":"https://openalex.org/C56372850","wikidata":"https://www.wikidata.org/wiki/Q1050404","display_name":"Sparse matrix","level":3,"score":0.5795000195503235},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.5608000159263611},{"id":"https://openalex.org/C70710897","wikidata":"https://www.wikidata.org/wiki/Q680081","display_name":"Separable space","level":2,"score":0.5533999800682068},{"id":"https://openalex.org/C21308566","wikidata":"https://www.wikidata.org/wiki/Q7169365","display_name":"Permutation (music)","level":2,"score":0.550000011920929},{"id":"https://openalex.org/C187834632","wikidata":"https://www.wikidata.org/wiki/Q188804","display_name":"Factorization","level":2,"score":0.516700029373169},{"id":"https://openalex.org/C155512373","wikidata":"https://www.wikidata.org/wiki/Q287450","display_name":"Residual","level":2,"score":0.4964999854564667},{"id":"https://openalex.org/C164226766","wikidata":"https://www.wikidata.org/wiki/Q7293202","display_name":"Rank (graph theory)","level":2,"score":0.4959999918937683},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.4909000098705292},{"id":"https://openalex.org/C124066611","wikidata":"https://www.wikidata.org/wiki/Q28684319","display_name":"Sparse approximation","level":2,"score":0.47999998927116394},{"id":"https://openalex.org/C130367717","wikidata":"https://www.wikidata.org/wiki/Q189791","display_name":"Diagonal","level":2,"score":0.478300005197525},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.47189998626708984},{"id":"https://openalex.org/C106487976","wikidata":"https://www.wikidata.org/wiki/Q685816","display_name":"Matrix (chemical analysis)","level":2,"score":0.4681999981403351},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.44290000200271606},{"id":"https://openalex.org/C77618280","wikidata":"https://www.wikidata.org/wiki/Q1155772","display_name":"Scheme (mathematics)","level":2,"score":0.42910000681877136},{"id":"https://openalex.org/C85817219","wikidata":"https://www.wikidata.org/wiki/Q884772","display_name":"Block matrix","level":3,"score":0.3758000135421753},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.31850001215934753},{"id":"https://openalex.org/C114614502","wikidata":"https://www.wikidata.org/wiki/Q76592","display_name":"Combinatorics","level":1,"score":0.27559998631477356},{"id":"https://openalex.org/C176809094","wikidata":"https://www.wikidata.org/wiki/Q15401496","display_name":"Traverse","level":2,"score":0.2743000090122223},{"id":"https://openalex.org/C2779304628","wikidata":"https://www.wikidata.org/wiki/Q3503480","display_name":"Face (sociological concept)","level":2,"score":0.2669999897480011},{"id":"https://openalex.org/C42355184","wikidata":"https://www.wikidata.org/wiki/Q1361088","display_name":"Matrix decomposition","level":3,"score":0.26649999618530273},{"id":"https://openalex.org/C88626702","wikidata":"https://www.wikidata.org/wiki/Q1128903","display_name":"Continuation","level":2,"score":0.262800008058548},{"id":"https://openalex.org/C78548338","wikidata":"https://www.wikidata.org/wiki/Q2493","display_name":"Data compression","level":2,"score":0.259799987077713}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3799830.3799837","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3799830.3799837","pdf_url":null,"source":null,"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 13th ACM IKDD International Conference on Data Science","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3799830.3799837","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3799830.3799837","pdf_url":null,"source":null,"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 13th ACM IKDD International Conference on Data Science","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":30,"referenced_works":["https://openalex.org/W1517701899","https://openalex.org/W1940947342","https://openalex.org/W1979511352","https://openalex.org/W1984201362","https://openalex.org/W2017056202","https://openalex.org/W2022532533","https://openalex.org/W2030963577","https://openalex.org/W2040740620","https://openalex.org/W2053107666","https://openalex.org/W2057777349","https://openalex.org/W2072118131","https://openalex.org/W2074093855","https://openalex.org/W2077866399","https://openalex.org/W2081385752","https://openalex.org/W2082222857","https://openalex.org/W2090560926","https://openalex.org/W2126566496","https://openalex.org/W2162947735","https://openalex.org/W2281635067","https://openalex.org/W3093647367","https://openalex.org/W3120468868","https://openalex.org/W4206749106","https://openalex.org/W4226466768","https://openalex.org/W4377088454","https://openalex.org/W4388191129","https://openalex.org/W4391111857","https://openalex.org/W4391179459","https://openalex.org/W4403693850","https://openalex.org/W4410335192","https://openalex.org/W7133227460"],"related_works":[],"abstract_inverted_index":{"Modern":[0],"large":[1],"language":[2],"models":[3],"(LLMs)":[4],"place":[5],"extraordinary":[6],"pressure":[7],"on":[8],"memory":[9],"and":[10,20,45,65,103,109],"compute":[11],"budgets,":[12],"making":[13],"principled":[14],"compression":[15],"indispensable":[16],"for":[17],"both":[18],"deployment":[19],"continued":[21],"training.":[22],"We":[23],"present":[24],"Hierarchical":[25],"Sparse":[26,51],"Plus":[27],"Low-Rank":[28],"(HSS)":[29,53],"compression,":[30],"a":[31,41,48,66,104],"two\u2013stage":[32],"scheme":[33],"that":[34],"(i)":[35],"removes":[36],"the":[37,57,78,81],"largest\u2010magnitude":[38],"weights":[39,76],"into":[40],"sparse":[42,102],"matrix":[43],"S":[44],"(ii)":[46],"applies":[47],"recursive":[49,62],"Hierarchically":[50],"Separable":[52],"low-rank":[54],"factorisation":[55],"to":[56,73,100],"dense":[58],"residual":[59],"matrix.":[60],"A":[61],"rank-reducing":[63],"strategy":[64],"reverse":[67],"Cuthill\u2013Mckee":[68],"(RCM)":[69],"permutation":[70],"are":[71,89],"introduced":[72],"align":[74],"high":[75],"towards":[77],"diagonal":[79],"with":[80,114],"block-diagonal":[82],"hierarchy,":[83],"maximising":[84],"off-diagonal":[85],"compressibility":[86],"(because":[87],"they":[88],"touched":[90],"only":[91],"once).":[92],"HSS":[93],"is":[94],"hardware-friendly:":[95],"its":[96],"matrix\u2013vector":[97],"multiply":[98],"reduces":[99],"one":[101],"sequence":[105],"of":[106],"thin-matrix":[107],"multiplications,":[108],"can":[110],"be":[111],"trained":[112],"end-to-end":[113],"standard":[115],"optimisers.":[116]},"counts_by_year":[],"updated_date":"2026-04-29T09:16:38.111599","created_date":"2026-04-25T00:00:00"}