{"id":"https://openalex.org/W4403679014","doi":"https://doi.org/10.48550/arxiv.2409.09927","title":"Towards Data Contamination Detection for Modern Large Language Models: Limitations, Inconsistencies, and Oracle Challenges","display_name":"Towards Data Contamination Detection for Modern Large Language Models: Limitations, Inconsistencies, and Oracle Challenges","publication_year":2024,"publication_date":"2024-09-16","ids":{"openalex":"https://openalex.org/W4403679014","doi":"https://doi.org/10.48550/arxiv.2409.09927"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2409.09927","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2409.09927","pdf_url":"https://arxiv.org/pdf/2409.09927","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2409.09927","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5072439636","display_name":"Vinay Samuel","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Samuel, Vinay","raw_affiliation_strings":["Carnegie Mellon University ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Carnegie Mellon University ,","institution_ids":["https://openalex.org/I74973139"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Zhou, Yue","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I39422238","display_name":"University of Illinois Chicago","ror":"https://ror.org/02mpq6x41","country_code":"US","type":"education","lineage":["https://openalex.org/I39422238"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhou, Yue","raw_affiliation_strings":["University of Illinois Chicago"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Illinois Chicago","institution_ids":["https://openalex.org/I39422238"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5109737659","display_name":"Henry Peng Zou","orcid":"https://orcid.org/0009-0003-5259-4998","observed_orcids":["https://orcid.org/0009-0003-5259-4998"]},"institutions":[{"id":"https://openalex.org/I39422238","display_name":"University of Illinois Chicago","ror":"https://ror.org/02mpq6x41","country_code":"US","type":"education","lineage":["https://openalex.org/I39422238"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zou, Henry Peng","raw_affiliation_strings":["University of Illinois Chicago"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Illinois Chicago","institution_ids":["https://openalex.org/I39422238"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":3,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.801800012588501,"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":0.801800012588501,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.1088000014424324,"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":0.023099999874830246,"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/large-language-models","display_name":"large language models","score":0.9670000076293945},{"id":"https://openalex.org/keywords/memorization","display_name":"memorization","score":0.6669999957084656},{"id":"https://openalex.org/keywords/llm-benchmark","display_name":"LLM benchmark","score":0.6840000152587891},{"id":"https://openalex.org/keywords/instruction-tuning","display_name":"instruction tuning","score":0.871999979019165}],"concepts":[{"id":"https://openalex.org/C55166926","wikidata":"https://www.wikidata.org/wiki/Q2892946","display_name":"Oracle","level":2,"score":0.6355677247047424},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6283044815063477},{"id":"https://openalex.org/C112570922","wikidata":"https://www.wikidata.org/wiki/Q60528603","display_name":"Contamination","level":2,"score":0.5516393184661865},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.4074922800064087},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.1325664222240448},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0689379870891571},{"id":"https://openalex.org/C18903297","wikidata":"https://www.wikidata.org/wiki/Q7150","display_name":"Ecology","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2409.09927","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2409.09927","pdf_url":"https://arxiv.org/pdf/2409.09927","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2409.09927","doi":"https://doi.org/10.48550/arxiv.2409.09927","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2409.09927","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2409.09927","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2409.09927","pdf_url":"https://arxiv.org/pdf/2409.09927","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4403679014.pdf","grobid_xml":"https://content.openalex.org/works/W4403679014.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7166495070","https://openalex.org/W4407806804","https://openalex.org/W4404088284","https://openalex.org/W4402670101","https://openalex.org/W4404783013","https://openalex.org/W4417287236","https://openalex.org/W4411120124","https://openalex.org/W7166876690","https://openalex.org/W7027667564","https://openalex.org/W4385965989"],"abstract_inverted_index":{"As":[0],"large":[1],"language":[2],"models":[3],"achieve":[4],"increasingly":[5],"impressive":[6],"results,":[7],"questions":[8],"arise":[9],"about":[10,44],"whether":[11],"such":[12],"performance":[13],"is":[14,161],"from":[15],"generalizability":[16],"or":[17],"mere":[18],"data":[19,23],"memorization.":[20],"Thus,":[21],"numerous":[22],"contamination":[24,53,70,79,118,132,141,157],"detection":[25,73,80,133,142],"methods":[26,102],"have":[27,103],"been":[28],"proposed.":[29],"However,":[30],"these":[31],"approaches":[32,81],"are":[33],"often":[34,90],"validated":[35],"with":[36,82,123],"traditional":[37],"benchmarks":[38],"and":[39,62,72,109,126,146,155],"early-stage":[40],"LLMs,":[41],"leaving":[42],"uncertainty":[43],"their":[45,107],"effectiveness":[46],"when":[47],"evaluating":[48],"state-of-the-art":[49,84],"LLMs":[50,85,145],"on":[51,153],"the":[52,138,147],"of":[54,67,140],"more":[55],"challenging":[56,88],"benchmarks.":[57],"To":[58],"address":[59],"this":[60],"gap":[61],"provide":[63],"a":[64],"dual":[65],"investigation":[66],"SOTA":[68,131],"LLM":[69,94],"status":[71],"method":[74],"robustness,":[75],"we":[76],"evaluate":[77],"five":[78],"four":[83],"across":[86],"eight":[87],"datasets":[89],"used":[91],"in":[92,106,116,143],"modern":[93],"evaluation.":[95,158],"Our":[96,159],"analysis":[97],"reveals":[98],"that":[99],"(1)":[100],"Current":[101],"non-trivial":[104],"limitations":[105],"assumptions":[108],"practical":[110],"applications;":[111],"(2)":[112],"Notable":[113],"difficulties":[114],"exist":[115],"detecting":[117],"introduced":[119],"during":[120],"instruction":[121],"fine-tuning":[122],"answer":[124],"augmentation;":[125],"(3)":[127],"Limited":[128],"consistencies":[129],"between":[130],"techniques.":[134],"These":[135],"findings":[136],"highlight":[137],"complexity":[139],"advanced":[144],"urgent":[148],"need":[149],"for":[150],"further":[151],"research":[152],"robust":[154],"generalizable":[156],"code":[160],"available":[162],"at":[163],"https://github.com/vsamuel2003/data-contamination.":[164]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":2}],"updated_date":"2026-10-28T10:33:24.094632","created_date":"2024-10-24T00:00:00"}