{"id":"https://openalex.org/W7101749573","doi":"https://doi.org/10.1145/3721462.3770776","title":"Leveraging Approximate Caching for Faster Retrieval-Augmented Generation","display_name":"Leveraging Approximate Caching for Faster Retrieval-Augmented Generation","publication_year":2025,"publication_date":"2025-12-08","ids":{"openalex":"https://openalex.org/W7101749573","doi":"https://doi.org/10.1145/3721462.3770776"},"language":null,"primary_location":{"id":"doi:10.1145/3721462.3770776","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3721462.3770776","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 26th International Middleware Conference","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Shai Bergman","orcid":"https://orcid.org/0000-0002-3484-1452"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Shai Bergman","raw_affiliation_strings":["Huawei, Zurich, Switzerland"],"affiliations":[{"raw_affiliation_string":"Huawei, Zurich, Switzerland","institution_ids":[]}]},{"author_position":"middle","author":{"id":null,"display_name":"Anne-Marie Kermarrec","orcid":"https://orcid.org/0000-0001-8187-724X"},"institutions":[{"id":"https://openalex.org/I5124864","display_name":"\u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne","ror":"https://ror.org/02s376052","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I5124864"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Anne-Marie Kermarrec","raw_affiliation_strings":["EPFL, Lausanne, Switzerland"],"affiliations":[{"raw_affiliation_string":"EPFL, Lausanne, Switzerland","institution_ids":["https://openalex.org/I5124864"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Diana Petrescu","orcid":"https://orcid.org/0009-0006-2229-235X"},"institutions":[{"id":"https://openalex.org/I5124864","display_name":"\u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne","ror":"https://ror.org/02s376052","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I5124864"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Diana Petrescu","raw_affiliation_strings":["EPFL, Lausanne, Switzerland"],"affiliations":[{"raw_affiliation_string":"EPFL, Lausanne, Switzerland","institution_ids":["https://openalex.org/I5124864"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Rafael Pires","orcid":"https://orcid.org/0000-0002-7826-1599"},"institutions":[{"id":"https://openalex.org/I5124864","display_name":"\u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne","ror":"https://ror.org/02s376052","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I5124864"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Rafael Pires","raw_affiliation_strings":["EPFL, Lausanne, Switzerland"],"affiliations":[{"raw_affiliation_string":"EPFL, Lausanne, Switzerland","institution_ids":["https://openalex.org/I5124864"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Mathis Randl","orcid":"https://orcid.org/0009-0003-6844-4695"},"institutions":[{"id":"https://openalex.org/I5124864","display_name":"\u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne","ror":"https://ror.org/02s376052","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I5124864"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Mathis Randl","raw_affiliation_strings":["EPFL, Lausanne, Switzerland"],"affiliations":[{"raw_affiliation_string":"EPFL, Lausanne, Switzerland","institution_ids":["https://openalex.org/I5124864"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Martijn de Vos","orcid":"https://orcid.org/0000-0003-4157-4847"},"institutions":[{"id":"https://openalex.org/I5124864","display_name":"\u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne","ror":"https://ror.org/02s376052","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I5124864"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Martijn de Vos","raw_affiliation_strings":["EPFL, Lausanne, Switzerland"],"affiliations":[{"raw_affiliation_string":"EPFL, Lausanne, Switzerland","institution_ids":["https://openalex.org/I5124864"]}]},{"author_position":"last","author":{"id":null,"display_name":"Ji Zhang","orcid":"https://orcid.org/0009-0008-6184-7279"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ji Zhang","raw_affiliation_strings":["Huawei, Zurich, Switzerland"],"affiliations":[{"raw_affiliation_string":"Huawei, Zurich, Switzerland","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":7,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":3.2508,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.94701626,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":96,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"340","last_page":"353"},"is_retracted":false,"is_paratext":false,"is_xpac":true,"primary_topic":{"id":"https://openalex.org/T10286","display_name":"Information Retrieval and Search Behavior","score":0.2305999994277954,"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"}},"topics":[{"id":"https://openalex.org/T10286","display_name":"Information Retrieval and Search Behavior","score":0.2305999994277954,"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"}},{"id":"https://openalex.org/T10627","display_name":"Advanced Image and Video Retrieval Techniques","score":0.13279999792575836,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.06530000269412994,"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/cache","display_name":"Cache","score":0.8069999814033508},{"id":"https://openalex.org/keywords/scheme","display_name":"Scheme (mathematics)","score":0.4959999918937683},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.48989999294281006},{"id":"https://openalex.org/keywords/cache-algorithms","display_name":"Cache algorithms","score":0.4108999967575073},{"id":"https://openalex.org/keywords/reliability","display_name":"Reliability (semiconductor)","score":0.4043999910354614},{"id":"https://openalex.org/keywords/smart-cache","display_name":"Smart Cache","score":0.3824000060558319},{"id":"https://openalex.org/keywords/cpu-cache","display_name":"CPU cache","score":0.3700000047683716},{"id":"https://openalex.org/keywords/workflow","display_name":"Workflow","score":0.3603000044822693}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8600999712944031},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.8069999814033508},{"id":"https://openalex.org/C77618280","wikidata":"https://www.wikidata.org/wiki/Q1155772","display_name":"Scheme (mathematics)","level":2,"score":0.4959999918937683},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.48989999294281006},{"id":"https://openalex.org/C38556500","wikidata":"https://www.wikidata.org/wiki/Q13404475","display_name":"Cache algorithms","level":4,"score":0.4108999967575073},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.4043999910354614},{"id":"https://openalex.org/C167713795","wikidata":"https://www.wikidata.org/wiki/Q352090","display_name":"Smart Cache","level":5,"score":0.3824000060558319},{"id":"https://openalex.org/C189783530","wikidata":"https://www.wikidata.org/wiki/Q352090","display_name":"CPU cache","level":3,"score":0.3700000047683716},{"id":"https://openalex.org/C177212765","wikidata":"https://www.wikidata.org/wiki/Q627335","display_name":"Workflow","level":2,"score":0.3603000044822693},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.34940001368522644},{"id":"https://openalex.org/C103278499","wikidata":"https://www.wikidata.org/wiki/Q254465","display_name":"Similarity (geometry)","level":3,"score":0.3474000096321106},{"id":"https://openalex.org/C2778476105","wikidata":"https://www.wikidata.org/wiki/Q628539","display_name":"Workload","level":2,"score":0.34549999237060547},{"id":"https://openalex.org/C99138194","wikidata":"https://www.wikidata.org/wiki/Q183427","display_name":"Hash function","level":2,"score":0.34290000796318054},{"id":"https://openalex.org/C77088390","wikidata":"https://www.wikidata.org/wiki/Q8513","display_name":"Database","level":1,"score":0.3125999867916107},{"id":"https://openalex.org/C81669768","wikidata":"https://www.wikidata.org/wiki/Q2359161","display_name":"Precision and recall","level":2,"score":0.2946999967098236},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.2915000021457672},{"id":"https://openalex.org/C2779808786","wikidata":"https://www.wikidata.org/wiki/Q6664603","display_name":"Locality","level":2,"score":0.2799000144004822},{"id":"https://openalex.org/C74270461","wikidata":"https://www.wikidata.org/wiki/Q1625299","display_name":"Locality-sensitive hashing","level":4,"score":0.27970001101493835},{"id":"https://openalex.org/C93996380","wikidata":"https://www.wikidata.org/wiki/Q44127","display_name":"Server","level":2,"score":0.2770000100135803},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.2750999927520752},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.27219998836517334},{"id":"https://openalex.org/C2777027219","wikidata":"https://www.wikidata.org/wiki/Q1284190","display_name":"Constant (computer programming)","level":2,"score":0.25440001487731934},{"id":"https://openalex.org/C150807984","wikidata":"https://www.wikidata.org/wiki/Q1992074","display_name":"Bit array","level":3,"score":0.25200000405311584}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3721462.3770776","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3721462.3770776","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 26th International Middleware Conference","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education","score":0.4523877501487732}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":18,"referenced_works":["https://openalex.org/W2005054045","https://openalex.org/W2050076590","https://openalex.org/W2124509324","https://openalex.org/W2126356880","https://openalex.org/W2806803384","https://openalex.org/W2963469388","https://openalex.org/W4285199301","https://openalex.org/W4313331827","https://openalex.org/W4388409299","https://openalex.org/W4388584589","https://openalex.org/W4398757454","https://openalex.org/W4400641571","https://openalex.org/W4402827393","https://openalex.org/W4403899386","https://openalex.org/W4403977136","https://openalex.org/W4404534210","https://openalex.org/W4405942816","https://openalex.org/W4415124086"],"related_works":[],"abstract_inverted_index":{"Retrieval-augmented":[0],"generation":[1],"(RAG)":[2],"improves":[3],"the":[4,19,47,73,103,147,151,162],"reliability":[5],"of":[6,57],"large":[7,29],"language":[8],"model":[9],"(LLM)":[10],"answers":[11],"by":[12,50,126],"integrating":[13],"external":[14],"knowledge.":[15],"However,":[16],"RAG":[17,48],"increases":[18],"end-to-end":[20],"inference":[21],"time":[22,148],"since":[23],"looking":[24],"for":[25,180],"relevant":[26],"documents":[27,66],"from":[28],"vector":[30,77],"databases":[31],"is":[32,174],"computationally":[33],"expensive.":[34],"To":[35,80],"address":[36],"this,":[37],"we":[38],"introduce":[39],"Proximity,":[40],"an":[41],"approximate":[42,172],"key-value":[43],"cache":[44,93,142,153,163],"that":[45,90,112,146,171],"optimizes":[46],"workflow":[49],"leveraging":[51],"similarities":[52],"in":[53,166],"user":[54],"queries.":[55],"Instead":[56],"treating":[58],"each":[59],"query":[60],"independently,":[61],"Proximity":[62,83,101,113,152],"reuses":[63],"previously":[64],"retrieved":[65],"when":[67],"similar":[68],"queries":[69],"appear,":[70],"substantially":[71,165],"reducing":[72],"reliance":[74],"on":[75],"expensive":[76],"database":[78,124,130],"lookups.":[79],"efficiently":[81],"scale,":[82],"employs":[84],"a":[85,119,175],"locality-sensitive":[86],"hashing":[87],"(LSH)":[88],"scheme":[89,117],"enables":[91],"fast":[92],"lookups":[94],"while":[95,128],"preserving":[96],"retrieval":[97],"accuracy.":[98,134],"We":[99,135],"evaluate":[100],"using":[102],"MMLU":[104],"and":[105,118,132,141,144,156,177],"MedRAG":[106,121],"question-answering":[107],"benchmarks.":[108],"Our":[109,168],"experiments":[110],"demonstrate":[111,170],"with":[114,137],"our":[115],"LSH":[116],"realistically-skewed":[120],"workload":[122],"reduces":[123],"calls":[125],"77.2%":[127],"maintaining":[129],"recall":[131],"test":[133],"experiment":[136],"different":[138],"similarity":[139],"tolerances":[140],"capacities,":[143],"show":[145],"spent":[149],"within":[150],"remains":[154],"low":[155],"constant":[157],"(4.8":[158],"\u03bcs)":[159],"even":[160],"as":[161],"grows":[164],"size.":[167],"results":[169],"caching":[173],"practical":[176],"effective":[178],"strategy":[179],"optimizing":[181],"RAG-based":[182],"systems.":[183]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-10-29T00:00:00"}