{"id":"https://openalex.org/W7161999610","doi":"https://doi.org/10.48550/arxiv.2605.21422","title":"PRISM: Preference-Aware Influence Function Based Data Selection Method for Efficient Fine-Tuning","display_name":"PRISM: Preference-Aware Influence Function Based Data Selection Method for Efficient Fine-Tuning","publication_year":2026,"publication_date":"2026-05-20","ids":{"openalex":"https://openalex.org/W7161999610","doi":"https://doi.org/10.48550/arxiv.2605.21422"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.21422","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.21422","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,"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.21422","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5039973913","display_name":"Qihao Lin","orcid":"https://orcid.org/0000-0002-1685-0187"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lin, Qihao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136716183","display_name":"Guanxu Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Guanxu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136638218","display_name":"Dongrui Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Dongrui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5136712283","display_name":"Jing Shao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shao, Jing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"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/T10743","display_name":"Software Testing and Debugging Techniques","score":0.10890000313520432,"subfield":{"id":"https://openalex.org/subfields/1712","display_name":"Software"},"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/T10743","display_name":"Software Testing and Debugging Techniques","score":0.10890000313520432,"subfield":{"id":"https://openalex.org/subfields/1712","display_name":"Software"},"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.09860000014305115,"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":0.0763000026345253,"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/weighting","display_name":"Weighting","score":0.6880000233650208},{"id":"https://openalex.org/keywords/selection","display_name":"Selection (genetic algorithm)","score":0.6617000102996826},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.5620999932289124},{"id":"https://openalex.org/keywords/relevance","display_name":"Relevance (law)","score":0.539900004863739},{"id":"https://openalex.org/keywords/preference","display_name":"Preference","score":0.505299985408783},{"id":"https://openalex.org/keywords/construct","display_name":"Construct (python library)","score":0.5040000081062317},{"id":"https://openalex.org/keywords/data-set","display_name":"Data set","score":0.43849998712539673},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.4253000020980835}],"concepts":[{"id":"https://openalex.org/C183115368","wikidata":"https://www.wikidata.org/wiki/Q856577","display_name":"Weighting","level":2,"score":0.6880000233650208},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.6617000102996826},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6274999976158142},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.5620999932289124},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.5547999739646912},{"id":"https://openalex.org/C158154518","wikidata":"https://www.wikidata.org/wiki/Q7310970","display_name":"Relevance (law)","level":2,"score":0.539900004863739},{"id":"https://openalex.org/C2781249084","wikidata":"https://www.wikidata.org/wiki/Q908656","display_name":"Preference","level":2,"score":0.505299985408783},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.5040000081062317},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4713999927043915},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.45159998536109924},{"id":"https://openalex.org/C58489278","wikidata":"https://www.wikidata.org/wiki/Q1172284","display_name":"Data set","level":2,"score":0.43849998712539673},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.4253000020980835},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.42239999771118164},{"id":"https://openalex.org/C67666897","wikidata":"https://www.wikidata.org/wiki/Q165896","display_name":"Prism","level":2,"score":0.3968999981880188},{"id":"https://openalex.org/C93959086","wikidata":"https://www.wikidata.org/wiki/Q6888345","display_name":"Model selection","level":2,"score":0.337799996137619},{"id":"https://openalex.org/C70136482","wikidata":"https://www.wikidata.org/wiki/Q13583781","display_name":"A-weighting","level":3,"score":0.3375000059604645},{"id":"https://openalex.org/C160920958","wikidata":"https://www.wikidata.org/wiki/Q7662746","display_name":"Synthetic data","level":2,"score":0.31859999895095825},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.29030001163482666},{"id":"https://openalex.org/C148043351","wikidata":"https://www.wikidata.org/wiki/Q4456944","display_name":"Current (fluid)","level":2,"score":0.2808000147342682},{"id":"https://openalex.org/C148483581","wikidata":"https://www.wikidata.org/wiki/Q446488","display_name":"Feature selection","level":2,"score":0.2662999927997589},{"id":"https://openalex.org/C16910744","wikidata":"https://www.wikidata.org/wiki/Q7705759","display_name":"Test data","level":2,"score":0.2554999887943268}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.21422","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.21422","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,"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":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.21422","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.21422","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,"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"},"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":{"As":[0],"LLMs":[1],"continue":[2],"to":[3,26,79,122,125,140,151,158],"scale":[4],"up,":[5],"improving":[6],"training":[7,24,51,137],"efficiency":[8],"heavily":[9],"relies":[10],"on":[11,54,58,143],"effective":[12],"data":[13,52,148,223],"utilization.":[14],"Data":[15,115],"selection":[16],"mitigates":[17],"this":[18,144],"issue":[19],"by":[20],"allocating":[21],"the":[22,32,73,89,156,219],"limited":[23],"budget":[25,149],"high-value":[27],"examples":[28,47,68,78,84,99,127],"that":[29,85,153,166,195,211],"optimally":[30],"facilitate":[31],"model's":[33,90],"target":[34,40,46,67,83,126,132,161,178,215],"behavior.":[35,162],"Most":[36],"existing":[37],"approaches":[38],"define":[39],"behavior":[41,92,216],"via":[42],"a":[43,110,130,171],"set":[44],"of":[45,76,214,221],"and":[48,103,128,146,191,203],"score":[49],"candidate":[50,136],"based":[53,114],"their":[55,141],"estimated":[56],"influence":[57,142],"these":[59],"samples.":[60],"However,":[61],"such":[62],"methods":[63],"uniformly":[64],"treat":[65],"all":[66],"as":[69,218],"equally":[70],"important,":[71],"ignoring":[72],"varying":[74],"relevance":[75],"individual":[77],"model":[80,120,157,189],"optimization.":[81],"Specifically,":[82],"align":[86],"closely":[87],"with":[88,181],"inherent":[91],"deliver":[93],"stronger":[94],"supervisory":[95],"signals,":[96],"whereas":[97],"discrepant":[98],"yield":[100],"only":[101],"weak":[102],"ineffective":[104],"local":[105],"guidance.":[106],"We":[107],"propose":[108],"PRISM,":[109],"Preference-aware":[111],"Influence":[112],"function":[113],"Selection":[116],"Method.":[117],"It":[118],"leverages":[119],"preference":[121,168],"assign":[123],"weights":[124],"builds":[129],"preference-aware":[131],"direction.":[133],"PRISM":[134,196],"evaluates":[135],"samples":[138,152],"according":[139],"direction,":[145],"prioritizes":[147],"allocation":[150],"effectively":[154],"drive":[155],"match":[159],"expected":[160],"Theoretical":[163],"analysis":[164],"verifies":[165],"weighted":[167],"construction":[169],"generates":[170],"superior":[172],"first-order":[173],"gradient":[174],"direction":[175],"for":[176],"boosting":[177],"preference,":[179],"compared":[180],"uniform":[182],"aggregation":[183],"strategies.":[184],"Extensive":[185],"experiments":[186],"covering":[187],"diverse":[188],"architectures":[190],"parameter":[192],"scales":[193],"demonstrate":[194],"achieves":[197],"better":[198],"performance":[199],"in":[200],"efficient":[201],"fine-tuning":[202,206],"safety-aligned":[204],"supervised":[205],"rectification.":[207],"The":[208],"results":[209],"validate":[210],"accurate":[212],"characterization":[213],"serves":[217],"core":[220],"cost-effective":[222],"selection.":[224]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-05-22T00:00:00"}