{"id":"https://openalex.org/W7133533176","doi":"https://doi.org/10.48550/arxiv.2603.03203","title":"No Memorization, No Detection: Output Distribution-Based Contamination Detection in Small Language Models","display_name":"No Memorization, No Detection: Output Distribution-Based Contamination Detection in Small Language Models","publication_year":2026,"publication_date":"2026-03-03","ids":{"openalex":"https://openalex.org/W7133533176","doi":"https://doi.org/10.48550/arxiv.2603.03203"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.03203","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.03203","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2603.03203","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5128057624","display_name":"Omer Sela","orcid":null},"institutions":[{"id":"https://openalex.org/I16391192","display_name":"Tel Aviv University","ror":"https://ror.org/04mhzgx49","country_code":"IL","type":"education","lineage":["https://openalex.org/I16391192"]}],"countries":["IL"],"is_corresponding":true,"raw_author_name":"Sela, Omer","raw_affiliation_strings":["Tel Aviv University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tel Aviv University","institution_ids":["https://openalex.org/I16391192"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":["https://openalex.org/A5128057624"],"corresponding_institution_ids":["https://openalex.org/I16391192"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.10649999976158142,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.10649999976158142,"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":0.09809999912977219,"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/T12072","display_name":"Machine Learning and Algorithms","score":0.08500000089406967,"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/memorization","display_name":"Memorization","score":0.659500002861023},{"id":"https://openalex.org/keywords/contamination","display_name":"Contamination","score":0.536300003528595},{"id":"https://openalex.org/keywords/code","display_name":"Code (set theory)","score":0.439300000667572},{"id":"https://openalex.org/keywords/ranging","display_name":"Ranging","score":0.35749998688697815},{"id":"https://openalex.org/keywords/source-code","display_name":"Source code","score":0.33390000462532043}],"concepts":[{"id":"https://openalex.org/C30038468","wikidata":"https://www.wikidata.org/wiki/Q4354775","display_name":"Memorization","level":2,"score":0.659500002861023},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6121000051498413},{"id":"https://openalex.org/C112570922","wikidata":"https://www.wikidata.org/wiki/Q60528603","display_name":"Contamination","level":2,"score":0.536300003528595},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.439300000667572},{"id":"https://openalex.org/C115051666","wikidata":"https://www.wikidata.org/wiki/Q6522493","display_name":"Ranging","level":2,"score":0.35749998688697815},{"id":"https://openalex.org/C43126263","wikidata":"https://www.wikidata.org/wiki/Q128751","display_name":"Source code","level":2,"score":0.33390000462532043},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.32499998807907104},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3246000111103058},{"id":"https://openalex.org/C62649853","wikidata":"https://www.wikidata.org/wiki/Q199687","display_name":"Remote sensing","level":1,"score":0.2703999876976013},{"id":"https://openalex.org/C12426560","wikidata":"https://www.wikidata.org/wiki/Q189569","display_name":"Basis (linear algebra)","level":2,"score":0.2667999863624573},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.25619998574256897},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.2508000135421753}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.03203","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.03203","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2603.03203","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.03203","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"CDD,":[0],"or":[1],"Contamination":[2],"Detection":[3],"via":[4],"output":[5],"Distribution,":[6],"identifies":[7],"data":[8,77],"contamination":[9,42,114],"by":[10,83],"measuring":[11],"the":[12,21,63,76],"peakedness":[13],"of":[14,65],"a":[15],"model's":[16],"sampled":[17],"outputs.":[18],"We":[19,86],"study":[20],"conditions":[22,66,100],"under":[23],"which":[24],"this":[25],"approach":[26,110],"succeeds":[27],"and":[28,47,81,93],"fails":[29],"on":[30,44,56],"small":[31,117],"language":[32,118],"models":[33],"ranging":[34],"from":[35],"70M":[36],"to":[37],"410M":[38],"parameters.":[39],"Using":[40],"controlled":[41],"experiments":[43],"GSM8K,":[45],"HumanEval,":[46],"MATH,":[48],"we":[49,67],"find":[50],"that":[51,88,107],"CDD's":[52,108],"effectiveness":[53],"depends":[54],"critically":[55],"whether":[57],"fine-tuning":[58],"produces":[59],"verbatim":[60],"memorization.":[61],"In":[62],"majority":[64],"test,":[68],"CDD":[69,97],"performs":[70],"at":[71,124],"chance":[72],"level":[73],"even":[74],"when":[75],"is":[78,111,122],"verifiably":[79],"contaminated":[80],"detectable":[82],"simpler":[84],"methods.":[85],"show":[87],"probability-based":[89],"methods,":[90],"specifically":[91],"perplexity":[92],"Min-k\\%":[94],"Prob,":[95],"outperform":[96],"in":[98,116],"all":[99],"where":[101],"any":[102],"method":[103],"exceeds":[104],"chance,":[105],"suggesting":[106],"peakedness-based":[109],"insufficient":[112],"for":[113],"detection":[115],"models.":[119],"Our":[120],"code":[121],"available":[123],"https://github.com/Sela-Omer/Contamination-Detection-Small-LM":[125]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-03-05T00:00:00"}