{"id":"https://openalex.org/W7128502470","doi":"https://doi.org/10.1016/j.future.2026.108423","title":"SpectraBench: A three-stage evolution framework for intelligent large language model evaluation","display_name":"SpectraBench: A three-stage evolution framework for intelligent large language model evaluation","publication_year":2026,"publication_date":"2026-02-10","ids":{"openalex":"https://openalex.org/W7128502470","doi":"https://doi.org/10.1016/j.future.2026.108423"},"language":"en","primary_location":{"id":"doi:10.1016/j.future.2026.108423","is_oa":true,"landing_page_url":"https://doi.org/10.1016/j.future.2026.108423","pdf_url":null,"source":{"id":"https://openalex.org/S186357190","display_name":"Future Generation Computer Systems","issn_l":"0167-739X","issn":["0167-739X","1872-7115"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320990","host_organization_name":"Elsevier BV","host_organization_lineage":["https://openalex.org/P4310320990"],"host_organization_lineage_names":["Elsevier BV"],"type":"journal"},"license":"cc-by-nc","license_id":"https://openalex.org/licenses/cc-by-nc","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Future Generation Computer Systems","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://doi.org/10.1016/j.future.2026.108423","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5125531220","display_name":"Gunwoo Lee","orcid":null},"institutions":[{"id":"https://openalex.org/I878022262","display_name":"Korea Institute of Science & Technology Information","ror":"https://ror.org/01k4yrm29","country_code":"KR","type":"facility","lineage":["https://openalex.org/I2801339556","https://openalex.org/I4210144908","https://openalex.org/I4387152098","https://openalex.org/I878022262"]}],"countries":["KR"],"is_corresponding":true,"raw_author_name":"Gunwoo Lee","raw_affiliation_strings":["Korea Institute of Science and Technology Information (KISTI), Large-scale AI Research Center, Daehak-ro 245 Yuseong-gu, Daejeon, 34141, Republic of Korea"],"raw_orcid":"https://orcid.org/0000-0002-5676-7556","affiliations":[{"raw_affiliation_string":"Korea Institute of Science and Technology Information (KISTI), Large-scale AI Research Center, Daehak-ro 245 Yuseong-gu, Daejeon, 34141, Republic of Korea","institution_ids":["https://openalex.org/I878022262"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125524232","display_name":"Rae-Young Jang","orcid":null},"institutions":[{"id":"https://openalex.org/I878022262","display_name":"Korea Institute of Science & Technology Information","ror":"https://ror.org/01k4yrm29","country_code":"KR","type":"facility","lineage":["https://openalex.org/I2801339556","https://openalex.org/I4210144908","https://openalex.org/I4387152098","https://openalex.org/I878022262"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Rae-Young Jang","raw_affiliation_strings":["Korea Institute of Science and Technology Information (KISTI), Large-scale AI Research Center, Daehak-ro 245 Yuseong-gu, Daejeon, 34141, Republic of Korea"],"raw_orcid":"https://orcid.org/0000-0001-9391-4028","affiliations":[{"raw_affiliation_string":"Korea Institute of Science and Technology Information (KISTI), Large-scale AI Research Center, Daehak-ro 245 Yuseong-gu, Daejeon, 34141, Republic of Korea","institution_ids":["https://openalex.org/I878022262"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125512255","display_name":"Sang-Hwan Gwak","orcid":null},"institutions":[{"id":"https://openalex.org/I878022262","display_name":"Korea Institute of Science & Technology Information","ror":"https://ror.org/01k4yrm29","country_code":"KR","type":"facility","lineage":["https://openalex.org/I2801339556","https://openalex.org/I4210144908","https://openalex.org/I4387152098","https://openalex.org/I878022262"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Sang-Hwan Gwak","raw_affiliation_strings":["Korea Institute of Science and Technology Information (KISTI), Large-scale AI Research Center, Daehak-ro 245 Yuseong-gu, Daejeon, 34141, Republic of Korea"],"raw_orcid":"https://orcid.org/0000-0001-9098-3001","affiliations":[{"raw_affiliation_string":"Korea Institute of Science and Technology Information (KISTI), Large-scale AI Research Center, Daehak-ro 245 Yuseong-gu, Daejeon, 34141, Republic of Korea","institution_ids":["https://openalex.org/I878022262"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5080763205","display_name":"Kyong-Ha Lee","orcid":"https://orcid.org/0000-0001-6929-0825"},"institutions":[{"id":"https://openalex.org/I878022262","display_name":"Korea Institute of Science & Technology Information","ror":"https://ror.org/01k4yrm29","country_code":"KR","type":"facility","lineage":["https://openalex.org/I2801339556","https://openalex.org/I4210144908","https://openalex.org/I4387152098","https://openalex.org/I878022262"]}],"countries":["KR"],"is_corresponding":true,"raw_author_name":"Kyong-Ha Lee","raw_affiliation_strings":["Korea Institute of Science and Technology Information (KISTI), Large-scale AI Research Center, Daehak-ro 245 Yuseong-gu, Daejeon, 34141, Republic of Korea"],"raw_orcid":"https://orcid.org/0000-0001-6929-0825","affiliations":[{"raw_affiliation_string":"Korea Institute of Science and Technology Information (KISTI), Large-scale AI Research Center, Daehak-ro 245 Yuseong-gu, Daejeon, 34141, Republic of Korea","institution_ids":["https://openalex.org/I878022262"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5055425748","display_name":"Ryong Lee","orcid":"https://orcid.org/0000-0001-5142-6106"},"institutions":[{"id":"https://openalex.org/I878022262","display_name":"Korea Institute of Science & Technology Information","ror":"https://ror.org/01k4yrm29","country_code":"KR","type":"facility","lineage":["https://openalex.org/I2801339556","https://openalex.org/I4210144908","https://openalex.org/I4387152098","https://openalex.org/I878022262"]}],"countries":["KR"],"is_corresponding":true,"raw_author_name":"Ryong Lee","raw_affiliation_strings":["Korea Institute of Science and Technology Information (KISTI), Large-scale AI Research Center, Daehak-ro 245 Yuseong-gu, Daejeon, 34141, Republic of Korea"],"raw_orcid":"https://orcid.org/0000-0001-5142-6106","affiliations":[{"raw_affiliation_string":"Korea Institute of Science and Technology Information (KISTI), Large-scale AI Research Center, Daehak-ro 245 Yuseong-gu, Daejeon, 34141, Republic of Korea","institution_ids":["https://openalex.org/I878022262"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5055425748","https://openalex.org/A5080763205","https://openalex.org/A5125531220"],"corresponding_institution_ids":["https://openalex.org/I878022262"],"apc_list":{"value":3340,"currency":"USD","value_usd":3340},"apc_paid":{"value":3340,"currency":"USD","value_usd":3340},"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.17374202,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"181","issue":null,"first_page":"108423","last_page":"108423"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.21240000426769257,"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/T10028","display_name":"Topic Modeling","score":0.21240000426769257,"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/T12380","display_name":"Authorship Attribution and Profiling","score":0.07980000227689743,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.05400000140070915,"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/language-model","display_name":"Language model","score":0.3630000054836273},{"id":"https://openalex.org/keywords/natural-language","display_name":"Natural language","score":0.2865999937057495},{"id":"https://openalex.org/keywords/language-understanding","display_name":"Language understanding","score":0.2639999985694885},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.25839999318122864},{"id":"https://openalex.org/keywords/constructed-language","display_name":"Constructed language","score":0.25619998574256897}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.9118000268936157},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.487199991941452},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3944999873638153},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3630000054836273},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.2865999937057495},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2784999907016754},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.267300009727478},{"id":"https://openalex.org/C2983448237","wikidata":"https://www.wikidata.org/wiki/Q1078276","display_name":"Language understanding","level":2,"score":0.2639999985694885},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.25839999318122864},{"id":"https://openalex.org/C94922259","wikidata":"https://www.wikidata.org/wiki/Q33215","display_name":"Constructed language","level":2,"score":0.25619998574256897},{"id":"https://openalex.org/C9652623","wikidata":"https://www.wikidata.org/wiki/Q190109","display_name":"Field (mathematics)","level":2,"score":0.2549000084400177},{"id":"https://openalex.org/C115903868","wikidata":"https://www.wikidata.org/wiki/Q80993","display_name":"Software engineering","level":1,"score":0.25459998846054077}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1016/j.future.2026.108423","is_oa":true,"landing_page_url":"https://doi.org/10.1016/j.future.2026.108423","pdf_url":null,"source":{"id":"https://openalex.org/S186357190","display_name":"Future Generation Computer Systems","issn_l":"0167-739X","issn":["0167-739X","1872-7115"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320990","host_organization_name":"Elsevier BV","host_organization_lineage":["https://openalex.org/P4310320990"],"host_organization_lineage_names":["Elsevier BV"],"type":"journal"},"license":"cc-by-nc","license_id":"https://openalex.org/licenses/cc-by-nc","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Future Generation Computer Systems","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1016/j.future.2026.108423","is_oa":true,"landing_page_url":"https://doi.org/10.1016/j.future.2026.108423","pdf_url":null,"source":{"id":"https://openalex.org/S186357190","display_name":"Future Generation Computer Systems","issn_l":"0167-739X","issn":["0167-739X","1872-7115"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320990","host_organization_name":"Elsevier BV","host_organization_lineage":["https://openalex.org/P4310320990"],"host_organization_lineage_names":["Elsevier BV"],"type":"journal"},"license":"cc-by-nc","license_id":"https://openalex.org/licenses/cc-by-nc","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Future Generation Computer Systems","raw_type":"journal-article"},"sustainable_development_goals":[{"score":0.4114803671836853,"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4"}],"awards":[{"id":"https://openalex.org/G4184913192","display_name":null,"funder_award_id":"K26L3M1C1","funder_id":"https://openalex.org/F4320322105","funder_display_name":"Korea Institute of Science and Technology Information"}],"funders":[{"id":"https://openalex.org/F4320322105","display_name":"Korea Institute of Science and Technology Information","ror":"https://ror.org/01k4yrm29"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":8,"referenced_works":["https://openalex.org/W1678356000","https://openalex.org/W2119821739","https://openalex.org/W2160293203","https://openalex.org/W2911964244","https://openalex.org/W3085139254","https://openalex.org/W3214897310","https://openalex.org/W4378189609","https://openalex.org/W4389519352"],"related_works":[],"abstract_inverted_index":{"Evaluating":[0],"Large":[1],"Language":[2],"Models":[3],"(LLMs)":[4],"is":[5,163],"one":[6],"of":[7],"the":[8,13,182],"most":[9],"resource-intensive":[10],"steps":[11],"in":[12,21,197],"AI":[14],"pipeline.":[15],"Conventional":[16],"sequential":[17],"execution":[18,143],"often":[19],"results":[20],"long":[22],"runtimes,":[23],"frequent":[24],"out-of-memory":[25],"(OOM)":[26],"failures,":[27],"and":[28,38,89,107,114,136,148,159,166,170,186,201],"poor":[29],"GPU":[30,36],"utilization.":[31],"Existing":[32],"frameworks":[33],"underuse":[34],"modern":[35],"capabilities":[37],"provide":[39],"limited":[40],"failure":[41],"prevention":[42],"for":[43,72,97],"large":[44],"models.":[45],"We":[46],"present":[47],"SpectraBench,":[48],"an":[49],"adaptive":[50],"scheduling":[51],"framework":[52,64],"that":[53,80],"optimizes":[54],"LLM":[55,191],"benchmarking":[56],"through":[57],"machine":[58],"learning":[59],"(ML)-driven":[60],"resource":[61,199],"orchestration.":[62],"The":[63],"follows":[65],"a":[66],"three-stage":[67],"process:":[68],"(1)":[69],"Foundation":[70],"Scheduling":[71,79,92,154],"cold-starts":[73],"using":[74,93],"heuristic":[75],"rules,":[76],"(2)":[77],"Hybrid":[78,153],"fuses":[81],"heuristics":[82],"with":[83,123,168],"ML":[84],"predictions":[85],"via":[86],"confidence-weighted":[87],"decisions,":[88],"(3)":[90],"Autonomous":[91],"Random":[94],"Forest":[95],"ensembles":[96],"automated":[98],"optimization.":[99],"SpectraBench":[100,139,162,193],"integrates":[101],"continuous":[102],"monitoring,":[103],"predictive":[104],"memory":[105,183],"management,":[106],"dynamic":[108],"parameter":[109],"tuning":[110],"to":[111,190],"prevent":[112],"failures":[113],"maximize":[115],"efficiency.":[116],"A":[117],"two-phase":[118],"optimization":[119],"combining":[120],"Convergence":[121],"Analysis":[122],"Threshold":[124],"Optimization":[125],"establishes":[126],"validated":[127],"transition":[128],"points":[129],"between":[130],"stages.":[131],"Across":[132],"13":[133],"LLMs":[134],"(0.5B\u201332B)":[135],"10":[137],"benchmarks,":[138],"achieves":[140],"10.3%":[141],"shorter":[142],"time,":[144],"11.8%":[145],"higher":[146],"throughput,":[147],"5.5%":[149],"lower":[150],"thermal":[151],"output.":[152],"consistently":[155],"outperforms":[156],"both":[157],"heuristic-only":[158],"ML-only":[160],"baselines.":[161],"production-ready,":[164],"modular,":[165],"compatible":[167],"single-":[169],"multi-GPU":[171],"systems,":[172],"enabling":[173],"seamless":[174],"integration":[175],"into":[176],"existing":[177],"evaluation":[178],"workflows.":[179],"By":[180],"addressing":[181],"pressure,":[184],"scheduling,":[185],"failure-prevention":[187],"challenges":[188],"specific":[189],"evaluation,":[192],"delivers":[194],"substantial":[195],"gains":[196],"performance,":[198],"efficiency,":[200],"reliability.":[202]},"counts_by_year":[],"updated_date":"2026-06-14T06:11:07.267592","created_date":"2026-02-11T00:00:00"}