{"id":"https://openalex.org/W4414359366","doi":"https://doi.org/10.24963/ijcai.2025/911","title":"Contamination Budget: Trade-offs Between Breadth, Depth and Difficulty","display_name":"Contamination Budget: Trade-offs Between Breadth, Depth and Difficulty","publication_year":2025,"publication_date":"2025-09-01","ids":{"openalex":"https://openalex.org/W4414359366","doi":"https://doi.org/10.24963/ijcai.2025/911"},"language":"en","primary_location":{"id":"doi:10.24963/ijcai.2025/911","is_oa":false,"landing_page_url":"https://doi.org/10.24963/ijcai.2025/911","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5092966916","display_name":"Behzad Mehrbakhsh","orcid":"https://orcid.org/0000-0001-9017-989X"},"institutions":[{"id":"https://openalex.org/I60053951","display_name":"Universitat Polit\u00e8cnica de Val\u00e8ncia","ror":"https://ror.org/01460j859","country_code":"ES","type":"education","lineage":["https://openalex.org/I60053951"]},{"id":"https://openalex.org/I4210131846","display_name":"Artificial Intelligence Research Institute","ror":"https://ror.org/03c0ach84","country_code":"ES","type":"facility","lineage":["https://openalex.org/I134820265","https://openalex.org/I4210131846"]}],"countries":["ES"],"is_corresponding":true,"raw_author_name":"Behzad Mehrbakhsh","raw_affiliation_strings":["UPV - Universitat Polit\u00e8cnica de Val\u00e8ncia","VRAIN - Valencian Research Institute for Artificial Intelligence"],"affiliations":[{"raw_affiliation_string":"UPV - Universitat Polit\u00e8cnica de Val\u00e8ncia","institution_ids":["https://openalex.org/I60053951"]},{"raw_affiliation_string":"VRAIN - Valencian Research Institute for Artificial Intelligence","institution_ids":["https://openalex.org/I4210131846"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5051725651","display_name":"Fernando Mart\u00ednez\u2010Plumed","orcid":"https://orcid.org/0000-0003-2902-6477"},"institutions":[{"id":"https://openalex.org/I60053951","display_name":"Universitat Polit\u00e8cnica de Val\u00e8ncia","ror":"https://ror.org/01460j859","country_code":"ES","type":"education","lineage":["https://openalex.org/I60053951"]},{"id":"https://openalex.org/I4210131846","display_name":"Artificial Intelligence Research Institute","ror":"https://ror.org/03c0ach84","country_code":"ES","type":"facility","lineage":["https://openalex.org/I134820265","https://openalex.org/I4210131846"]}],"countries":["ES"],"is_corresponding":false,"raw_author_name":"Fernando Mart\u00ednez-Plumed","raw_affiliation_strings":["UPV - Universitat Polit\u00e8cnica de Val\u00e8ncia","VRAIN - Valencian Research Institute for Artificial Intelligence"],"affiliations":[{"raw_affiliation_string":"UPV - Universitat Polit\u00e8cnica de Val\u00e8ncia","institution_ids":["https://openalex.org/I60053951"]},{"raw_affiliation_string":"VRAIN - Valencian Research Institute for Artificial Intelligence","institution_ids":["https://openalex.org/I4210131846"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5029864546","display_name":"Jos\u00e9 Hern\u00e1ndez\u2010Orallo","orcid":"https://orcid.org/0000-0001-9746-7632"},"institutions":[{"id":"https://openalex.org/I60053951","display_name":"Universitat Polit\u00e8cnica de Val\u00e8ncia","ror":"https://ror.org/01460j859","country_code":"ES","type":"education","lineage":["https://openalex.org/I60053951"]},{"id":"https://openalex.org/I4210131846","display_name":"Artificial Intelligence Research Institute","ror":"https://ror.org/03c0ach84","country_code":"ES","type":"facility","lineage":["https://openalex.org/I134820265","https://openalex.org/I4210131846"]}],"countries":["ES"],"is_corresponding":false,"raw_author_name":"Jos\u00e9 Hern\u00e1ndez-Orallo","raw_affiliation_strings":["UPV - Universitat Polit\u00e8cnica de Val\u00e8ncia","VRAIN - Valencian Research Institute for Artificial Intelligence"],"affiliations":[{"raw_affiliation_string":"UPV - Universitat Polit\u00e8cnica de Val\u00e8ncia","institution_ids":["https://openalex.org/I60053951"]},{"raw_affiliation_string":"VRAIN - Valencian Research Institute for Artificial Intelligence","institution_ids":["https://openalex.org/I4210131846"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5092966916"],"corresponding_institution_ids":["https://openalex.org/I4210131846","https://openalex.org/I60053951"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.13937434,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"8195","last_page":"8203"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.22390000522136688,"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":0.22390000522136688,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.11710000038146973,"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":0.10480000078678131,"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/task","display_name":"Task (project management)","score":0.5985000133514404},{"id":"https://openalex.org/keywords/test","display_name":"Test (biology)","score":0.5121999979019165},{"id":"https://openalex.org/keywords/phenomenon","display_name":"Phenomenon","score":0.48739999532699585},{"id":"https://openalex.org/keywords/intervention","display_name":"Intervention (counseling)","score":0.3855000138282776},{"id":"https://openalex.org/keywords/contamination","display_name":"Contamination","score":0.37209999561309814}],"concepts":[{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.5985000133514404},{"id":"https://openalex.org/C2777267654","wikidata":"https://www.wikidata.org/wiki/Q3519023","display_name":"Test (biology)","level":2,"score":0.5121999979019165},{"id":"https://openalex.org/C50335755","wikidata":"https://www.wikidata.org/wiki/Q483247","display_name":"Phenomenon","level":2,"score":0.48739999532699585},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.4657999873161316},{"id":"https://openalex.org/C2780665704","wikidata":"https://www.wikidata.org/wiki/Q959298","display_name":"Intervention (counseling)","level":2,"score":0.3855000138282776},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.37540000677108765},{"id":"https://openalex.org/C112570922","wikidata":"https://www.wikidata.org/wiki/Q60528603","display_name":"Contamination","level":2,"score":0.37209999561309814},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3529999852180481},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.32760000228881836},{"id":"https://openalex.org/C109359841","wikidata":"https://www.wikidata.org/wiki/Q728944","display_name":"Inclusion (mineral)","level":2,"score":0.31200000643730164},{"id":"https://openalex.org/C112930515","wikidata":"https://www.wikidata.org/wiki/Q4389547","display_name":"Risk analysis (engineering)","level":1,"score":0.29440000653266907},{"id":"https://openalex.org/C2776436953","wikidata":"https://www.wikidata.org/wiki/Q5163215","display_name":"Consistency (knowledge bases)","level":2,"score":0.27950000762939453},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2732999920845032}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.24963/ijcai.2025/911","is_oa":false,"landing_page_url":"https://doi.org/10.24963/ijcai.2025/911","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence","raw_type":"proceedings-article"}],"best_oa_location":null,"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":{"Contamination":[0],"in":[1,21,54,65,75,155],"large":[2],"language":[3],"models":[4],"(LLMs),":[5],"and":[6,24,105,115,168],"machine":[7],"learning":[8],"more":[9,126],"broadly,":[10],"refers":[11],"to":[12,94,128,163],"the":[13,66,76,79,86,89,106,109,158,166,169],"inclusion":[14],"of":[15,68,78,88,102,108,157],"equal":[16],"--or":[17],"very":[18],"similar--":[19],"examples":[20,93],"both":[22],"training":[23],"test":[25,33],"sets.":[26],"This":[27,146],"phenomenon":[28],"usually":[29,64],"translates":[30],"into":[31],"better":[32,52],"performance.":[34],"Here":[35],"we":[36,117],"explore":[37],"when":[38],"this":[39],"contamination":[40],"is":[41],"performed":[42],"intentionally,":[43],"for":[44],"purposes":[45],"that":[46,148],"can":[47],"be":[48,95,129,151],"malicious":[49],"(e.g.,":[50,58],"get":[51],"scores":[53],"evaluations)":[55],"or":[56],"benevolent":[57],"fix":[59],"some":[60,119,133],"mistakes).":[61],"These":[62],"interventions,":[63],"form":[67],"fine-tuning":[69,80],"memorisations,":[70],"come":[71],"with":[72],"a":[73],"budget":[74,159],"size":[77],"dataset.":[81],"Several":[82],"trade-offs":[83,149],"appear":[84],"between":[85],"breadth":[87],"intervention":[90],"(how":[91,99],"many":[92,100],"memorised),":[96],"its":[97],"depth":[98,127,138],"repetitions":[101],"each":[103],"example)":[104],"difficulty":[107,171],"examples.":[110],"By":[111],"studying":[112],"several":[113],"LLMs":[114],"datasets,":[116],"observe":[118],"monotonic":[120],"behaviour":[121],"(more":[122],"difficult":[123],"items":[124],"require":[125],"`fixed')":[130],"but":[131,160],"also":[132,161],"non-monotonic":[134],"phenomena":[135],"(very":[136],"high":[137],"levels":[139],"have":[140],"negative":[141],"effects":[142],"on":[143],"non-contaminated":[144],"examples).":[145],"suggests":[147],"should":[150],"found":[152],"not":[153],"only":[154],"terms":[156],"according":[162],"model":[164],"specifics,":[165],"task":[167],"item":[170],"at":[172],"hand.":[173]},"counts_by_year":[],"updated_date":"2026-03-07T16:01:11.037858","created_date":"2025-10-10T00:00:00"}