{"id":"https://openalex.org/W4412888345","doi":"https://doi.org/10.18653/v1/2025.findings-acl.576","title":"User Behavior Prediction as a Generic, Robust, Scalable, and Low-Cost Evaluation Strategy for Estimating Generalization in LLMs","display_name":"User Behavior Prediction as a Generic, Robust, Scalable, and Low-Cost Evaluation Strategy for Estimating Generalization in LLMs","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412888345","doi":"https://doi.org/10.18653/v1/2025.findings-acl.576"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.findings-acl.576","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.576","pdf_url":"https://aclanthology.org/2025.findings-acl.576.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2025","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.findings-acl.576.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5112913577","display_name":"Sougata Saha","orcid":null},"institutions":[{"id":"https://openalex.org/I4210164862","display_name":"Artificial Intelligence in Medicine (Canada)","ror":"https://ror.org/05p590m36","country_code":"CA","type":"company","lineage":["https://openalex.org/I4210164862"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Sougata Saha","raw_affiliation_strings":["University of Artificial Intelligence ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Artificial Intelligence ,","institution_ids":["https://openalex.org/I4210164862"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5008944385","display_name":"Monojit Choudhury","orcid":"https://orcid.org/0000-0001-7473-7839"},"institutions":[{"id":"https://openalex.org/I4210164862","display_name":"Artificial Intelligence in Medicine (Canada)","ror":"https://ror.org/05p590m36","country_code":"CA","type":"company","lineage":["https://openalex.org/I4210164862"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Monojit Choudhury","raw_affiliation_strings":["University of Artificial Intelligence ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Artificial Intelligence ,","institution_ids":["https://openalex.org/I4210164862"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I4210164862"],"apc_list":null,"apc_paid":null,"fwci":1.9246,"has_fulltext":true,"cited_by_count":1,"citation_normalized_percentile":{"value":0.84965587,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":97},"biblio":{"volume":null,"issue":null,"first_page":"11047","last_page":"11065"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10538","display_name":"Data Mining Algorithms and Applications","score":0.9404000043869019,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T10538","display_name":"Data Mining Algorithms and Applications","score":0.9404000043869019,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T12761","display_name":"Data Stream Mining Techniques","score":0.9182000160217285,"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/generalization","display_name":"Generalization","score":0.6821282505989075},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.674787163734436},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6487084627151489},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3702136278152466},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.35350972414016724},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.10951021313667297},{"id":"https://openalex.org/keywords/database","display_name":"Database","score":0.09681046009063721}],"concepts":[{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.6821282505989075},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.674787163734436},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6487084627151489},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3702136278152466},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.35350972414016724},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.10951021313667297},{"id":"https://openalex.org/C77088390","wikidata":"https://www.wikidata.org/wiki/Q8513","display_name":"Database","level":1,"score":0.09681046009063721},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.findings-acl.576","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.576","pdf_url":"https://aclanthology.org/2025.findings-acl.576.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2025","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.findings-acl.576","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.576","pdf_url":"https://aclanthology.org/2025.findings-acl.576.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2025","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412888345.pdf","grobid_xml":"https://content.openalex.org/works/W4412888345.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W2961085424","https://openalex.org/W4306674287","https://openalex.org/W4387369504","https://openalex.org/W4394896187","https://openalex.org/W3170094116","https://openalex.org/W4386462264","https://openalex.org/W3107602296","https://openalex.org/W4364306694","https://openalex.org/W4312192474","https://openalex.org/W4283697347"],"abstract_inverted_index":{"Measuring":[0],"the":[1],"generalization":[2],"ability":[3],"of":[4,64],"Large":[5],"Language":[6],"Models":[7],"(LLMs)":[8],"is":[9],"challenging":[10],"due":[11],"to":[12],"data":[13],"contamination.As":[14],"models":[15,108],"grow":[16],"and":[17,23,38,71,81,86,93,104],"computation":[18],"becomes":[19],"cheaper,":[20],"ensuring":[21],"tasks":[22,40],"test":[24,82],"cases":[25],"are":[26,41,49],"unseen":[27],"during":[28],"training":[29],"phases":[30],"will":[31],"become":[32],"nearly":[33],"impossible.We":[34],"argue":[35],"that":[36],"knowledge-retrieval":[37],"reasoning":[39],"not":[42,50],"ideal":[43],"for":[44,52,78,90,112],"measuring":[45],"generalization,":[46],"as":[47,66],"LLMs":[48],"trained":[51],"specific":[53],"tasks.Instead,":[54],"we":[55],"propose":[56],"user":[57],"behavior":[58],"prediction,":[59],"also":[60],"a":[61,67,75],"key":[62],"aspect":[63],"personalization,":[65],"theoretically":[68],"sound,":[69],"scalable,":[70],"robust":[72],"alternative.We":[73],"introduce":[74],"novel":[76],"framework":[77],"this":[79],"approach":[80],"it":[83],"on":[84],"movie":[85],"music":[87],"recommendation":[88],"datasets":[89],"GPT-4o,":[91],"GPT-4o-mini,":[92],"Llama-3.1-8B-Instruct.Results":[94],"align":[95],"with":[96],"our":[97],"framework's":[98],"predictions,":[99],"showing":[100],"GPT-4o":[101],"outperforms":[102],"GPT-4o-mini":[103],"Llama,":[105],"though":[106],"all":[107],"have":[109],"much":[110],"room":[111],"improvement,":[113],"especially":[114],"Llama.":[115]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}