{"id":"https://openalex.org/W7138176033","doi":"https://doi.org/10.1609/aaai.v40i27.39476","title":"AIM: Manifold-based Data Filtering for Representation Finetuning","display_name":"AIM: Manifold-based Data Filtering for Representation Finetuning","publication_year":2026,"publication_date":"2026-03-14","ids":{"openalex":"https://openalex.org/W7138176033","doi":"https://doi.org/10.1609/aaai.v40i27.39476"},"language":null,"primary_location":{"id":"doi:10.1609/aaai.v40i27.39476","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i27.39476","pdf_url":null,"source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"diamond","oa_url":"https://doi.org/10.1609/aaai.v40i27.39476","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5129663933","display_name":"Qing Li","orcid":null},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Qing Li","raw_affiliation_strings":["Shanghai Jiao Tong University"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129652115","display_name":"Qibin Zheng","orcid":null},"institutions":[{"id":"https://openalex.org/I4210096250","display_name":"Beijing Institute of Big Data Research","ror":"https://ror.org/00s1sz824","country_code":"CN","type":"facility","lineage":["https://openalex.org/I20231570","https://openalex.org/I37796252","https://openalex.org/I4210096250"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Qibin Zheng","raw_affiliation_strings":["Advanced Institute of Big Data,Beijing"],"affiliations":[{"raw_affiliation_string":"Advanced Institute of Big Data,Beijing","institution_ids":["https://openalex.org/I4210096250"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129735829","display_name":"Yi Liu","orcid":null},"institutions":[{"id":"https://openalex.org/I4210096250","display_name":"Beijing Institute of Big Data Research","ror":"https://ror.org/00s1sz824","country_code":"CN","type":"facility","lineage":["https://openalex.org/I20231570","https://openalex.org/I37796252","https://openalex.org/I4210096250"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yi Liu","raw_affiliation_strings":["Advanced Institute of Big Data,Beijing"],"affiliations":[{"raw_affiliation_string":"Advanced Institute of Big Data,Beijing","institution_ids":["https://openalex.org/I4210096250"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5129675644","display_name":"Xingchun Diao","orcid":null},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xingchun Diao","raw_affiliation_strings":["Shanghai Jiao Tong University"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University","institution_ids":["https://openalex.org/I183067930"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5129663933"],"corresponding_institution_ids":["https://openalex.org/I183067930"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.39079229,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"40","issue":"27","first_page":"23096","last_page":"23104"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11273","display_name":"Advanced Graph Neural Networks","score":0.26089999079704285,"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/T11273","display_name":"Advanced Graph Neural Networks","score":0.26089999079704285,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.14350000023841858,"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/T10028","display_name":"Topic Modeling","score":0.1128000020980835,"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/leverage","display_name":"Leverage (statistics)","score":0.6506999731063843},{"id":"https://openalex.org/keywords/rendering","display_name":"Rendering (computer graphics)","score":0.5152000188827515},{"id":"https://openalex.org/keywords/interpretability","display_name":"Interpretability","score":0.4909000098705292},{"id":"https://openalex.org/keywords/generalization","display_name":"Generalization","score":0.4884999990463257},{"id":"https://openalex.org/keywords/representation","display_name":"Representation (politics)","score":0.46869999170303345},{"id":"https://openalex.org/keywords/consistency","display_name":"Consistency (knowledge bases)","score":0.4643999934196472},{"id":"https://openalex.org/keywords/training-set","display_name":"Training set","score":0.44200000166893005},{"id":"https://openalex.org/keywords/deep-neural-networks","display_name":"Deep neural networks","score":0.38989999890327454},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.3734999895095825},{"id":"https://openalex.org/keywords/external-data-representation","display_name":"External Data Representation","score":0.37229999899864197}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7200999855995178},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.6506999731063843},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6172999739646912},{"id":"https://openalex.org/C205711294","wikidata":"https://www.wikidata.org/wiki/Q176953","display_name":"Rendering (computer graphics)","level":2,"score":0.5152000188827515},{"id":"https://openalex.org/C2781067378","wikidata":"https://www.wikidata.org/wiki/Q17027399","display_name":"Interpretability","level":2,"score":0.4909000098705292},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.4884999990463257},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.46869999170303345},{"id":"https://openalex.org/C2776436953","wikidata":"https://www.wikidata.org/wiki/Q5163215","display_name":"Consistency (knowledge bases)","level":2,"score":0.4643999934196472},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4627000093460083},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.44200000166893005},{"id":"https://openalex.org/C2984842247","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep neural networks","level":3,"score":0.38989999890327454},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.3734999895095825},{"id":"https://openalex.org/C116409475","wikidata":"https://www.wikidata.org/wiki/Q1385056","display_name":"External Data Representation","level":2,"score":0.37229999899864197},{"id":"https://openalex.org/C106131492","wikidata":"https://www.wikidata.org/wiki/Q3072260","display_name":"Filter (signal processing)","level":2,"score":0.36559998989105225},{"id":"https://openalex.org/C529865628","wikidata":"https://www.wikidata.org/wiki/Q1790740","display_name":"Manifold (fluid mechanics)","level":2,"score":0.36399999260902405},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.3594000041484833},{"id":"https://openalex.org/C95623464","wikidata":"https://www.wikidata.org/wiki/Q1096149","display_name":"Classifier (UML)","level":2,"score":0.32350000739097595},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.32190001010894775},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.3149999976158142},{"id":"https://openalex.org/C2776145971","wikidata":"https://www.wikidata.org/wiki/Q30673951","display_name":"Labeled data","level":2,"score":0.31200000643730164},{"id":"https://openalex.org/C160920958","wikidata":"https://www.wikidata.org/wiki/Q7662746","display_name":"Synthetic data","level":2,"score":0.31150001287460327},{"id":"https://openalex.org/C24756922","wikidata":"https://www.wikidata.org/wiki/Q1757694","display_name":"Data quality","level":3,"score":0.30469998717308044},{"id":"https://openalex.org/C136520226","wikidata":"https://www.wikidata.org/wiki/Q302814","display_name":"Geometric data analysis","level":2,"score":0.29989999532699585},{"id":"https://openalex.org/C56435381","wikidata":"https://www.wikidata.org/wiki/Q1196371","display_name":"Geometric transformation","level":3,"score":0.29409998655319214},{"id":"https://openalex.org/C162319229","wikidata":"https://www.wikidata.org/wiki/Q175263","display_name":"Data structure","level":2,"score":0.27480000257492065},{"id":"https://openalex.org/C2781020372","wikidata":"https://www.wikidata.org/wiki/Q533093","display_name":"On the fly","level":2,"score":0.26589998602867126},{"id":"https://openalex.org/C151876577","wikidata":"https://www.wikidata.org/wiki/Q7049464","display_name":"Nonlinear dimensionality reduction","level":3,"score":0.25839999318122864},{"id":"https://openalex.org/C204241405","wikidata":"https://www.wikidata.org/wiki/Q461499","display_name":"Transformation (genetics)","level":3,"score":0.2531999945640564},{"id":"https://openalex.org/C32990609","wikidata":"https://www.wikidata.org/wiki/Q306542","display_name":"Transformation geometry","level":2,"score":0.2513999938964844},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.25060001015663147}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1609/aaai.v40i27.39476","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i27.39476","pdf_url":null,"source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1609/aaai.v40i27.39476","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i27.39476","pdf_url":null,"source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Representation":[0],"Finetuning":[1],"(ReFT)":[2],"has":[3],"recently":[4],"emerged":[5],"as":[6,110],"an":[7],"efficient":[8],"paradigm":[9],"for":[10,132],"adapting":[11],"pretrained":[12],"language":[13],"models":[14],"by":[15,56,107,139],"editing":[16],"hidden":[17,62],"representations":[18,63],"rather":[19],"than":[20],"model":[21,187],"weights.":[22],"However,":[23],"our":[24],"preliminary":[25],"experiments":[26,164],"reveal":[27],"that":[28,60,74,90],"ReFT":[29,78],"is":[30],"notably":[31],"more":[32],"sensitive":[33],"to":[34,39,45,150],"training":[35,137],"data":[36,129,182],"quality":[37],"compared":[38],"traditional":[40],"parameter-efficient":[41],"finetuning":[42],"methods,":[43],"particularly":[44],"samples":[46,138],"with":[47,148],"incorrect":[48],"labels,":[49],"which":[50,134],"can":[51],"severely":[52],"degrade":[53],"performance.":[54],"Inspired":[55],"prior":[57],"work":[58],"demonstrating":[59],"the":[61,91,103,141,173],"of":[64,144,175],"generalizable":[65],"neural":[66],"networks":[67],"exhibit":[68],"low-dimensional":[69,100],"manifold":[70,154],"structures,":[71],"we":[72,118],"hypothesize":[73],"effective":[75],"generalization":[76],"in":[77],"requires":[79],"geometrically":[80],"structured":[81],"transformations":[82,96,105],"between":[83],"pre-":[84],"and":[85,168],"post-intervention":[86],"representations.":[87],"This":[88],"implies":[89],"intervention":[92,146],"vectors":[93,147],"representing":[94],"these":[95],"should":[97],"form":[98],"a":[99,123,127,151],"manifold,":[101],"rendering":[102],"inconsistent":[104],"induced":[106],"label":[108],"noise":[109],"detectable":[111],"geometric":[112,142],"outliers.":[113],"To":[114],"leverage":[115],"this":[116],"insight,":[117],"introduce":[119],"Aligning":[120],"Interventions":[121],"on":[122,160,165],"learned":[124],"Manifold":[125],"(AIM),":[126],"representation-based":[128],"filtering":[130],"method":[131],"ReFT,":[133],"identifies":[135],"high-quality":[136],"measuring":[140],"consistency":[143],"their":[145],"respect":[149],"robust":[152],"reference":[153],"derived":[155],"via":[156],"principal":[157],"component":[158],"analysis":[159],"trusted":[161],"data.":[162],"Extensive":[163],"both":[166],"commonsense":[167],"arithmetic":[169],"reasoning":[170],"tasks":[171],"confirm":[172],"effectiveness":[174],"AIM,":[176],"showing":[177],"consistent":[178],"improvements":[179],"over":[180],"strong":[181],"selection":[183],"baselines":[184],"across":[185],"multiple":[186],"scales.":[188]},"counts_by_year":[],"updated_date":"2026-03-20T20:47:17.329874","created_date":"2026-03-18T00:00:00"}