{"id":"https://openalex.org/W7203629809","doi":"https://doi.org/10.48550/arxiv.2608.13866","title":"Geometric Filtering of LLM-Generated Samples for Few-Shot Text Classification","display_name":"Geometric Filtering of LLM-Generated Samples for Few-Shot Text Classification","publication_year":2026,"publication_date":"2026-08-14","ids":{"openalex":"https://openalex.org/W7203629809","doi":"https://doi.org/10.48550/arxiv.2608.13866"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.13866","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.13866","pdf_url":"https://arxiv.org/pdf/2608.13866","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2608.13866","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5147336309","display_name":"Benjam\u00edn Schindler","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I21067949","display_name":"Adolfo Ib\u00e1\u00f1ez University","ror":"https://ror.org/0326knt82","country_code":"CL","type":"education","lineage":["https://openalex.org/I21067949"]}],"countries":["CL"],"is_corresponding":false,"raw_author_name":"Schindler, Benjam\u00edn","raw_affiliation_strings":["Faculty of Engineering and Sciences Universidad Adolfo Ib\u00e1\u00f1ez Millennium Nucleus for Social Data Science (SODAS) Santiago , Chile"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Faculty of Engineering and Sciences Universidad Adolfo Ib\u00e1\u00f1ez Millennium Nucleus for Social Data Science (SODAS) Santiago , Chile","institution_ids":["https://openalex.org/I21067949"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5136495021","display_name":"Gonzalo A. Ruz","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I21067949","display_name":"Adolfo Ib\u00e1\u00f1ez University","ror":"https://ror.org/0326knt82","country_code":"CL","type":"education","lineage":["https://openalex.org/I21067949"]}],"countries":["CL"],"is_corresponding":false,"raw_author_name":"Ruz, Gonzalo A.","raw_affiliation_strings":["Faculty of Engineering and Sciences Universidad Adolfo Ib\u00e1\u00f1ez Millennium Nucleus for Social Data Science (SODAS) Millennium Nucleus in Data Science for Plant Resilience (PhytoLearning) Santiago , Chile"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Faculty of Engineering and Sciences Universidad Adolfo Ib\u00e1\u00f1ez Millennium Nucleus for Social Data Science (SODAS) Millennium Nucleus in Data Science for Plant Resilience (PhytoLearning) Santiago , Chile","institution_ids":["https://openalex.org/I21067949"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I21067949"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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/T10028","display_name":"Topic Modeling","score":0.8032000064849854,"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":0.8032000064849854,"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/T11652","display_name":"Imbalanced Data Classification Techniques","score":0.09730000048875809,"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":0.042100001126527786,"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/geometric-filtering","display_name":"geometric filtering","score":0.6850000023841858},{"id":"https://openalex.org/keywords/few-shot-text-classification","display_name":"few-shot text classification","score":0.9369999766349792},{"id":"https://openalex.org/keywords/sentence-embeddings","display_name":"sentence embeddings","score":0.5730000138282776},{"id":"https://openalex.org/keywords/sample-weighting","display_name":"sample weighting","score":0.6100000143051147},{"id":"https://openalex.org/keywords/data-augmentation","display_name":"data augmentation","score":0.5360000133514404},{"id":"https://openalex.org/keywords/named-entity-recognition","display_name":"named entity recognition","score":0.9200000166893005}],"concepts":[{"id":"https://openalex.org/C41608201","wikidata":"https://www.wikidata.org/wiki/Q980509","display_name":"Embedding","level":2,"score":0.7027000188827515},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6660000085830688},{"id":"https://openalex.org/C95623464","wikidata":"https://www.wikidata.org/wiki/Q1096149","display_name":"Classifier (UML)","level":2,"score":0.6546000242233276},{"id":"https://openalex.org/C183115368","wikidata":"https://www.wikidata.org/wiki/Q856577","display_name":"Weighting","level":2,"score":0.644599974155426},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.5830000042915344},{"id":"https://openalex.org/C106131492","wikidata":"https://www.wikidata.org/wiki/Q3072260","display_name":"Filter (signal processing)","level":2,"score":0.5799999833106995},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5442000031471252},{"id":"https://openalex.org/C129782007","wikidata":"https://www.wikidata.org/wiki/Q162886","display_name":"Euclidean geometry","level":2,"score":0.5004000067710876},{"id":"https://openalex.org/C186450821","wikidata":"https://www.wikidata.org/wiki/Q17295","display_name":"Euclidean space","level":2,"score":0.49140000343322754},{"id":"https://openalex.org/C2777212361","wikidata":"https://www.wikidata.org/wiki/Q5127848","display_name":"Class (philosophy)","level":2,"score":0.47859999537467957},{"id":"https://openalex.org/C100279318","wikidata":"https://www.wikidata.org/wiki/Q467440","display_name":"Sample space","level":2,"score":0.43689998984336853},{"id":"https://openalex.org/C198531522","wikidata":"https://www.wikidata.org/wiki/Q485146","display_name":"Sample (material)","level":2,"score":0.42570000886917114},{"id":"https://openalex.org/C120174047","wikidata":"https://www.wikidata.org/wiki/Q847073","display_name":"Euclidean distance","level":2,"score":0.39640000462532043},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3653999865055084},{"id":"https://openalex.org/C2778572836","wikidata":"https://www.wikidata.org/wiki/Q380933","display_name":"Space (punctuation)","level":2,"score":0.30309998989105225},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.3003000020980835},{"id":"https://openalex.org/C2777530160","wikidata":"https://www.wikidata.org/wiki/Q41796","display_name":"Sentence","level":2,"score":0.2978000044822693},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.28119999170303345},{"id":"https://openalex.org/C160633673","wikidata":"https://www.wikidata.org/wiki/Q355198","display_name":"Pixel","level":2,"score":0.26759999990463257},{"id":"https://openalex.org/C21080849","wikidata":"https://www.wikidata.org/wiki/Q13611879","display_name":"Data point","level":2,"score":0.2639999985694885},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.25929999351501465}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.13866","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.13866","pdf_url":"https://arxiv.org/pdf/2608.13866","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2608.13866","doi":"https://doi.org/10.48550/arxiv.2608.13866","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.13866","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,"listed_in":[],"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"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2608.13866","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.13866","pdf_url":"https://arxiv.org/pdf/2608.13866","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G8167787340","display_name":"CENTER FOR MATHEMATICAL MODELING (CMM)","funder_award_id":"FB210005","funder_id":"https://openalex.org/F4320331146","funder_display_name":"Agencia Nacional de Investigaci\u00f3n y Desarrollo"},{"id":"https://openalex.org/G8964081475","display_name":"LEARNING THRESHOLD NETWORKS FOR SOCIAL AND BIOLOGICAL APPLICATIONS","funder_award_id":"1230315","funder_id":"https://openalex.org/F4320331146","funder_display_name":"Agencia Nacional de Investigaci\u00f3n y Desarrollo"}],"funders":[{"id":"https://openalex.org/F4320331146","display_name":"Agencia Nacional de Investigaci\u00f3n y Desarrollo","ror":"https://ror.org/05btm5233"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7203629809.pdf","grobid_xml":"https://content.openalex.org/works/W7203629809.grobid-xml"},"referenced_works_count":3,"referenced_works":["https://openalex.org/W4406983345","https://openalex.org/W4417298407","https://openalex.org/W7213430141"],"related_works":["https://openalex.org/W6929513424","https://openalex.org/W4406528970","https://openalex.org/W4403571372","https://openalex.org/W7203472590","https://openalex.org/W4410636854","https://openalex.org/W6966970433","https://openalex.org/W7132832057","https://openalex.org/W4404340889","https://openalex.org/W4389518772","https://openalex.org/W4416034012"],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2],"(LLMs)":[3],"can":[4],"generate":[5],"synthetic":[6],"training":[7],"data":[8],"for":[9,77],"text":[10],"classification,":[11],"but":[12],"the":[13,27,136],"quality":[14],"of":[15,26],"generated":[16],"samples":[17],"is":[18,123,134],"heterogeneous:":[19],"some":[20],"fall":[21],"in":[22,33,57],"correct":[23],"class":[24,55],"regions":[25],"embedding":[28,60],"space":[29],"while":[30],"others":[31],"land":[32],"peripheral":[34],"or":[35],"cross-class":[36],"zones.":[37],"We":[38],"propose":[39],"a":[40,58],"geometric":[41],"filtering":[42],"framework":[43],"that":[44,135],"evaluates":[45],"each":[46],"LLM-generated":[47],"sample":[48,75],"by":[49],"its":[50],"Euclidean":[51],"distance":[52],"to":[53,111],"real":[54],"examples":[56],"sentence":[59],"space,":[61],"selecting":[62],"only":[63],"geometrically":[64],"consistent":[65],"candidates.":[66],"A":[67,131],"soft":[68],"weighting":[69],"mechanism":[70],"transforms":[71],"filter":[72,120,139],"scores":[73],"into":[74],"weights":[76],"classifier":[78],"training.":[79],"Evaluated":[80],"across":[81,125],"13":[82],"datasets,":[83],"5":[84,126],"classifiers,":[85],"10":[86],"augmentation":[87],"methods,":[88],"and":[89,122],"over":[90,100],"6,700":[91],"configurations,":[92],"our":[93],"method":[94],"achieves":[95],"+2.61":[96],"percentage":[97],"points":[98],"(pp)":[99],"SMOTE":[101],"($p<0.0001$,":[102],"Cohen's":[103],"$d=0.95$,":[104],"88.9%":[105],"win":[106,117],"rate).":[107],"The":[108],"approach":[109],"generalizes":[110],"named":[112],"entity":[113],"recognition":[114],"(+9.26pp,":[115],"100%":[116],"rate)":[118],"without":[119],"modification,":[121],"robust":[124],"LLMs":[127],"from":[128],"4":[129],"providers.":[130],"key":[132],"finding":[133],"simplest":[137],"distance-based":[138],"consistently":[140],"outperforms":[141],"complex":[142],"multi-criteria":[143],"alternatives.":[144]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-08-18T00:00:00"}