{"id":"https://openalex.org/W4412945366","doi":"https://doi.org/10.18653/v1/2025.acl-long.656","title":"MMLU-CF: A Contamination-free Multi-task Language Understanding Benchmark","display_name":"MMLU-CF: A Contamination-free Multi-task Language Understanding Benchmark","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412945366","doi":"https://doi.org/10.18653/v1/2025.acl-long.656"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.acl-long.656","doi":"https://doi.org/10.18653/v1/2025.acl-long.656","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.656","pdf_url":"https://aclanthology.org/2025.acl-long.656.pdf","source":{"id":"https://openalex.org/S4393915740","display_name":"Proceedings of the conference - Association for Computational Linguistics. Meeting","issn_l":"0736-587X","issn":["0736-587X"],"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":["norway-1"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"bronze","oa_url":"https://aclanthology.org/2025.acl-long.656.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5053043601","display_name":"Qihao Zhao","orcid":"https://orcid.org/0000-0002-2344-2771","observed_orcids":["https://orcid.org/0000-0002-2344-2771"]},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Qihao Zhao","raw_affiliation_strings":["Microsoft Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Research","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114112585","display_name":"Yangyu Huang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yangyu Huang","raw_affiliation_strings":["Microsoft Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Research","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5031963349","display_name":"Tengchao Lv","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tengchao Lv","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101581993","display_name":"Lei Cui","orcid":"https://orcid.org/0000-0002-1527-2215","observed_orcids":["https://orcid.org/0000-0002-1527-2215"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lei Cui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5059272550","display_name":"Qi Sun","orcid":"https://orcid.org/0009-0005-0146-0724","observed_orcids":["https://orcid.org/0009-0005-0146-0724"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qinzheng Sun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5052035563","display_name":"Shaoguang Mao","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Shaoguang Mao","raw_affiliation_strings":["Microsoft Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Research","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100638344","display_name":"Xin Zhang","orcid":"https://orcid.org/0000-0002-5413-515X","observed_orcids":["https://orcid.org/0000-0002-5413-515X"]},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Xin Zhang","raw_affiliation_strings":["Microsoft Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Research","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Ying Xin","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ying Xin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Qiufeng Yin","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Qiufeng Yin","raw_affiliation_strings":["Microsoft Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Research","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5016920988","display_name":"Shan Li","orcid":"https://orcid.org/0000-0001-6825-0346","observed_orcids":["https://orcid.org/0000-0001-6825-0346"]},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Scarlett Li","raw_affiliation_strings":["Microsoft Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Research","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5014662947","display_name":"Furu Wei","orcid":"https://orcid.org/0000-0002-7810-5852","observed_orcids":["https://orcid.org/0000-0002-7810-5852"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Furu Wei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":10.3566,"has_fulltext":true,"cited_by_count":11,"citation_normalized_percentile":{"value":0.9867165,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":97,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"13371","last_page":"13391"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.7577999830245972,"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":0.7577999830245972,"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":0.1873999983072281,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.013399999588727951,"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/benchmark","display_name":"benchmark","score":0.5460000038146973},{"id":"https://openalex.org/keywords/multi-choice-question-answering","display_name":"multi-choice question answering","score":0.48500001430511475},{"id":"https://openalex.org/keywords/language-model-evaluation","display_name":"language model evaluation","score":0.3190000057220459}],"concepts":[{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.7149855494499207},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6823941469192505},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.5736293792724609},{"id":"https://openalex.org/C112570922","wikidata":"https://www.wikidata.org/wiki/Q60528603","display_name":"Contamination","level":2,"score":0.5647755861282349},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.33727750182151794},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.12823399901390076},{"id":"https://openalex.org/C127313418","wikidata":"https://www.wikidata.org/wiki/Q1069","display_name":"Geology","level":0,"score":0.11969342827796936},{"id":"https://openalex.org/C201995342","wikidata":"https://www.wikidata.org/wiki/Q682496","display_name":"Systems engineering","level":1,"score":0.09809795022010803},{"id":"https://openalex.org/C18903297","wikidata":"https://www.wikidata.org/wiki/Q7150","display_name":"Ecology","level":1,"score":0.0},{"id":"https://openalex.org/C13280743","wikidata":"https://www.wikidata.org/wiki/Q131089","display_name":"Geodesy","level":1,"score":0.0},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.18653/v1/2025.acl-long.656","doi":"https://doi.org/10.18653/v1/2025.acl-long.656","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.656","pdf_url":"https://aclanthology.org/2025.acl-long.656.pdf","source":{"id":"https://openalex.org/S4393915740","display_name":"Proceedings of the conference - Association for Computational Linguistics. Meeting","issn_l":"0736-587X","issn":["0736-587X"],"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":["norway-1"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2412.15194","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2412.15194","pdf_url":"https://arxiv.org/pdf/2412.15194","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2412.15194","doi":"https://doi.org/10.48550/arxiv.2412.15194","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2412.15194","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,"listed_in":[],"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"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.acl-long.656","doi":"https://doi.org/10.18653/v1/2025.acl-long.656","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.656","pdf_url":"https://aclanthology.org/2025.acl-long.656.pdf","source":{"id":"https://openalex.org/S4393915740","display_name":"Proceedings of the conference - Association for Computational Linguistics. Meeting","issn_l":"0736-587X","issn":["0736-587X"],"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":["norway-1"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412945366.pdf","grobid_xml":"https://content.openalex.org/works/W4412945366.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4388788574","https://openalex.org/W7159678828","https://openalex.org/W4388555916","https://openalex.org/W4399455140","https://openalex.org/W4412888759","https://openalex.org/W4404089351","https://openalex.org/W4398229687","https://openalex.org/W4404783013","https://openalex.org/W4416034719","https://openalex.org/W7138355761"],"abstract_inverted_index":{"Multiple-choice":[0],"question":[1],"(MCQ)":[2],"datasets":[3],"like":[4],"Massive":[5],"Multitask":[6],"Language":[7],"Understanding":[8],"(MMLU)":[9],"are":[10],"widely":[11],"used":[12],"to":[13,44,131,142,157],"evaluate":[14],"the":[15,26,33,64,114,136,155,158,169,177],"commonsense,":[16],"understanding,":[17],"and":[18,32,81,101,118,124,145,186],"problem-solving":[19],"abilities":[20],"of":[21,29,36,74,96,172,179],"large":[22],"language":[23],"models":[24],"(LLMs).However,":[25],"open-source":[27],"nature":[28],"these":[30],"benchmarks":[31],"broad":[34],"sources":[35],"training":[37],"data":[38,83,87,110],"for":[39],"LLMs":[40,153],"have":[41],"inevitably":[42],"led":[43],"benchmark":[45,67,115],"contamination,":[46,88,111],"which":[47,70],"is":[48,139],"studied":[49],"in":[50,56,182],"our":[51,180],"contamination":[52],"evaluation":[53,188],"experiment,":[54],"resulting":[55],"unreliable":[57],"evaluation.To":[58],"alleviate":[59],"this":[60],"issue,":[61],"we":[62,89,112],"propose":[63],"contamination-free":[65],"MCQ":[66],"called":[68],"MMLU-CF,":[69],"reassesses":[71],"LLMs'":[72,163],"understanding":[73],"world":[75],"knowledge":[76],"by":[77],"averting":[78],"both":[79],"unintentional":[80,86],"deliberate":[82,109],"contamination.To":[84],"mitigate":[85],"source":[90],"questions":[91],"from":[92],"a":[93,184],"broader":[94],"domain":[95],"over":[97,150],"200":[98],"billion":[99],"webpages":[100],"apply":[102],"three":[103],"specifically":[104],"designed":[105],"decontamination":[106],"rules.To":[107],"prevent":[108],"divide":[113],"into":[116],"validation":[117,137],"test":[119,127],"sets":[120],"with":[121],"similar":[122],"difficulty":[123],"subject":[125],"distributions.The":[126],"set":[128,138],"remains":[129],"closed-source":[130],"ensure":[132],"reliable":[133],"results,":[134],"while":[135],"publicly":[140],"available":[141],"promote":[143],"transparency":[144],"facilitate":[146],"independent":[147],"evaluation.We":[148],"evaluated":[149],"40":[151],"mainstream":[152],"on":[154],"MMLU-CF.Compared":[156],"original":[159],"MMLU,":[160],"not":[161],"only":[162],"performances":[164],"significantly":[165],"dropped":[166],"but":[167],"also":[168],"performance":[170],"rankings":[171],"them":[173],"changed":[174],"considerably.This":[175],"indicates":[176],"effectiveness":[178],"approach":[181],"establishing":[183],"contaminationfree":[185],"fairer":[187],"standard.":[189]},"counts_by_year":[{"year":2026,"cited_by_count":7},{"year":2025,"cited_by_count":4}],"updated_date":"2026-10-30T12:39:25.321706","created_date":"2025-10-10T00:00:00"}