{"id":"https://openalex.org/W4416037157","doi":"https://doi.org/10.18653/v1/2025.emnlp-industry.186","title":"Efficiency-Effectiveness Reranking FLOPs for LLM-based Rerankers","display_name":"Efficiency-Effectiveness Reranking FLOPs for LLM-based Rerankers","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4416037157","doi":"https://doi.org/10.18653/v1/2025.emnlp-industry.186"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2025.emnlp-industry.186","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.emnlp-industry.186","pdf_url":"https://aclanthology.org/2025.emnlp-industry.186.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":"Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing: Industry Track","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.emnlp-industry.186.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5120309175","display_name":"Zhiyuan Peng","orcid":null},"institutions":[{"id":"https://openalex.org/I16269868","display_name":"Santa Clara University","ror":"https://ror.org/03ypqe447","country_code":"US","type":"education","lineage":["https://openalex.org/I16269868"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhiyuan Peng","raw_affiliation_strings":["Santa Clara University ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Santa Clara University ,","institution_ids":["https://openalex.org/I16269868"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5120309176","display_name":"Ting-ruen Wei","orcid":null},"institutions":[{"id":"https://openalex.org/I16269868","display_name":"Santa Clara University","ror":"https://ror.org/03ypqe447","country_code":"US","type":"education","lineage":["https://openalex.org/I16269868"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ting-Ruen Wei","raw_affiliation_strings":["Santa Clara University ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Santa Clara University ,","institution_ids":["https://openalex.org/I16269868"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5120309177","display_name":"Tingyu Song","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tingyu Song","raw_affiliation_strings":["Independent Researcher ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Independent Researcher ,","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5120309178","display_name":"Yilun Zhao","orcid":null},"institutions":[{"id":"https://openalex.org/I32971472","display_name":"Yale University","ror":"https://ror.org/03v76x132","country_code":"US","type":"education","lineage":["https://openalex.org/I32971472"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yilun Zhao","raw_affiliation_strings":["Yale University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Yale University","institution_ids":["https://openalex.org/I32971472"]}]}],"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":0.38239329,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"2782","last_page":"2791"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10363","display_name":"Low-power high-performance VLSI design","score":0.06459999829530716,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10363","display_name":"Low-power high-performance VLSI design","score":0.06459999829530716,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T14189","display_name":"Intravenous Infusion Technology and Safety","score":0.04800000041723251,"subfield":{"id":"https://openalex.org/subfields/2204","display_name":"Biomedical Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10808","display_name":"Electric and Hybrid Vehicle Technologies","score":0.039900001138448715,"subfield":{"id":"https://openalex.org/subfields/2203","display_name":"Automotive Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/flops","display_name":"FLOPS","score":0.6754000186920166},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.3197000026702881},{"id":"https://openalex.org/keywords/sequence","display_name":"Sequence (biology)","score":0.26570001244544983},{"id":"https://openalex.org/keywords/identification","display_name":"Identification (biology)","score":0.2651999890804291},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.24650000035762787}],"concepts":[{"id":"https://openalex.org/C3826847","wikidata":"https://www.wikidata.org/wiki/Q188768","display_name":"FLOPS","level":2,"score":0.6754000186920166},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5985000133514404},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.3197000026702881},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2797999978065491},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.26570001244544983},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.2651999890804291},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.24650000035762787},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.24529999494552612},{"id":"https://openalex.org/C2779664074","wikidata":"https://www.wikidata.org/wiki/Q3518405","display_name":"Terminal (telecommunication)","level":2,"score":0.21780000627040863},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.2143000066280365}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.emnlp-industry.186","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.emnlp-industry.186","pdf_url":"https://aclanthology.org/2025.emnlp-industry.186.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":"Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing: Industry Track","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.emnlp-industry.186","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.emnlp-industry.186","pdf_url":"https://aclanthology.org/2025.emnlp-industry.186.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":"Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing: Industry Track","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4416037157.pdf","grobid_xml":"https://content.openalex.org/works/W4416037157.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"Language":[1],"Models":[2],"(LLMs)":[3],"have":[4],"recently":[5],"been":[6],"applied":[7],"to":[8,66,74,139,160,179],"reranking":[9],"tasks":[10],"in":[11],"information":[12],"retrieval,":[13],"achieving":[14],"strong":[15],"performance.However,":[16],"their":[17],"high":[18],"computational":[19],"demands":[20],"often":[21,64],"hinder":[22],"practical":[23],"deployment.Existing":[24],"studies":[25],"evaluate":[26,161],"the":[27,38,78,81,130,141,153,172,180,183],"efficiency":[28],"of":[29,40,80,143,165,182],"LLM-based":[30,94,145,166],"rerankers":[31,167],"using":[32],"proxy":[33],"metrics":[34,49,98],"such":[35],"as":[36],"latency,":[37],"number":[39],"forward":[41],"passes,":[42],"input":[43],"tokens,":[44],"and":[45,53,63,76,115,175],"output":[46],"tokens.However,":[47],"these":[48],"depend":[50],"on":[51,152],"hardware":[52],"running-time":[54],"choices":[55],"(e.g.,":[56,106],"parallel":[57],"or":[58,108],"not,":[59],"batch":[60],"size,":[61,70],"etc),":[62],"fail":[65],"account":[67],"for":[68,93],"model":[69],"making":[71],"it":[72],"difficult":[73],"interpret":[75],"obscuring":[77],"evaluation":[79],"efficiency-effectiveness":[82],"tradeoff.To":[83],"address":[84],"this":[85,177],"issue,":[86],"we":[87,156],"propose":[88],"E":[89],"2":[90],"R-FLOPS":[91],"1":[92],"rerankers:":[95],"RPP":[96],"(ranking":[97],"per":[99,113,118,127],"PetaFLOP),":[100,119],"measuring":[101,120],"how":[102,121],"much":[103],"ranking":[104],"quality":[105],"NDCG":[107],"MRR)":[109],"a":[110,162],"method":[111],"achieves":[112],"PetaFLOP,":[114],"QPP":[116],"(queries":[117],"many":[122],"queries":[123],"can":[124],"be":[125],"processed":[126],"PetaFLOP.Accompanied":[128],"by":[129],"new":[131],"metrics,":[132,155],"an":[133,144],"interpretable":[134],"FLOPs":[135,142],"estimator":[136],"is":[137],"developed":[138],"estimate":[140],"reranker":[146],"even":[147],"without":[148],"running":[149],"any":[150],"experiments.Based":[151],"proposed":[154],"conduct":[157],"comprehensive":[158],"experiments":[159],"wide":[163],"range":[164],"with":[168],"different":[169],"architectures,":[170],"studying":[171],"efficiencyeffectiveness":[173],"trade-off":[174],"bringing":[176],"issue":[178],"attention":[181],"research":[184],"community.":[185]},"counts_by_year":[],"updated_date":"2026-08-23T07:36:19.812096","created_date":"2025-11-08T00:00:00"}