{"id":"https://openalex.org/W7160905881","doi":"https://doi.org/10.48550/arxiv.2605.10075","title":"Active Testing of Large Language Models via Approximate Neyman Allocation","display_name":"Active Testing of Large Language Models via Approximate Neyman Allocation","publication_year":2026,"publication_date":"2026-05-11","ids":{"openalex":"https://openalex.org/W7160905881","doi":"https://doi.org/10.48550/arxiv.2605.10075"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.10075","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.10075","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.10075","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5135921358","display_name":"Zeli Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Zeli","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135931058","display_name":"Jiancheng Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Jiancheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135984500","display_name":"Cong Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Cong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5135930023","display_name":"Yinglun Zhu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhu, Yinglun","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/T10028","display_name":"Topic Modeling","score":0.311599999666214,"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":0.311599999666214,"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/T12072","display_name":"Machine Learning and Algorithms","score":0.1559000015258789,"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":0.05920000001788139,"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/bottleneck","display_name":"Bottleneck","score":0.5791000127792358},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.5558000206947327},{"id":"https://openalex.org/keywords/entropy","display_name":"Entropy (arrow of time)","score":0.46050000190734863},{"id":"https://openalex.org/keywords/reduction","display_name":"Reduction (mathematics)","score":0.4544000029563904},{"id":"https://openalex.org/keywords/principle-of-maximum-entropy","display_name":"Principle of maximum entropy","score":0.4287000000476837},{"id":"https://openalex.org/keywords/range","display_name":"Range (aeronautics)","score":0.4122999906539917},{"id":"https://openalex.org/keywords/generative-grammar","display_name":"Generative grammar","score":0.41119998693466187}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6940000057220459},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.5791000127792358},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5724999904632568},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.5558000206947327},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4821000099182129},{"id":"https://openalex.org/C106301342","wikidata":"https://www.wikidata.org/wiki/Q4117933","display_name":"Entropy (arrow of time)","level":2,"score":0.46050000190734863},{"id":"https://openalex.org/C111335779","wikidata":"https://www.wikidata.org/wiki/Q3454686","display_name":"Reduction (mathematics)","level":2,"score":0.4544000029563904},{"id":"https://openalex.org/C9679016","wikidata":"https://www.wikidata.org/wiki/Q1417473","display_name":"Principle of maximum entropy","level":2,"score":0.4287000000476837},{"id":"https://openalex.org/C204323151","wikidata":"https://www.wikidata.org/wiki/Q905424","display_name":"Range (aeronautics)","level":2,"score":0.4122999906539917},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.41119998693466187},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.3822999894618988},{"id":"https://openalex.org/C167981619","wikidata":"https://www.wikidata.org/wiki/Q1685498","display_name":"Cross entropy","level":3,"score":0.336899995803833},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.335999995470047},{"id":"https://openalex.org/C52740198","wikidata":"https://www.wikidata.org/wiki/Q1539564","display_name":"Importance sampling","level":3,"score":0.3181999921798706},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.3019999861717224},{"id":"https://openalex.org/C77967617","wikidata":"https://www.wikidata.org/wiki/Q4677561","display_name":"Active learning (machine learning)","level":2,"score":0.2775000035762787},{"id":"https://openalex.org/C3018395757","wikidata":"https://www.wikidata.org/wiki/Q1379672","display_name":"Evaluation methods","level":2,"score":0.2775000035762787},{"id":"https://openalex.org/C2780009758","wikidata":"https://www.wikidata.org/wiki/Q6804172","display_name":"Measure (data warehouse)","level":2,"score":0.2653999924659729},{"id":"https://openalex.org/C60008888","wikidata":"https://www.wikidata.org/wiki/Q6031013","display_name":"Information bottleneck method","level":3,"score":0.2648000121116638}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.10075","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.10075","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.10075","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.10075","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":"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"},"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":{"Large":[0],"language":[1,116],"models":[2,95],"(LLMs)":[3],"require":[4],"reliable":[5],"evaluation":[6,13,40,53,63,99],"from":[7,55,93,111],"pre-training":[8],"to":[9,46,85,96,139],"test-time":[10],"scaling,":[11],"making":[12],"a":[14,56,79,121],"recurring":[15],"rather":[16],"than":[17],"one-off":[18],"cost.":[19],"As":[20],"model":[21,125],"scales":[22],"grow":[23],"and":[24,34,71,101,117,120,133,146],"target":[25,69],"tasks":[26],"increasingly":[27],"demand":[28],"expert":[29],"annotators,":[30],"both":[31],"the":[32,52,62,98],"compute":[33],"labeling":[35],"costs":[36],"needed":[37],"for":[38],"each":[39],"rise":[41],"rapidly.":[42],"Active":[43],"testing":[44,82],"aims":[45],"alleviate":[47],"this":[48],"bottleneck":[49],"by":[50],"approximating":[51],"result":[54],"small":[57],"but":[58],"informative":[59],"subset":[60],"of":[61,123,149],"pool.":[64],"However,":[65],"existing":[66],"approaches":[67],"primarily":[68],"classification":[70],"break":[72],"down":[73],"on":[74,108,131],"generative":[75,86],"tasks.":[76,87],"We":[77],"introduce":[78],"novel":[80],"active":[81],"algorithm":[83],"tailored":[84],"Our":[88],"method":[89,128],"leverages":[90],"semantic":[91],"entropy":[92],"surrogate":[94],"stratify":[97],"pool":[100],"then":[102],"conducts":[103],"approximate":[104],"Neyman":[105],"allocation":[106],"based":[107],"signals":[109],"extracted":[110],"these":[112],"surrogates.":[113],"Across":[114],"multiple":[115],"multimodal":[118],"benchmarks":[119],"range":[122],"surrogate-target":[124],"pairs,":[126],"our":[127],"significantly":[129],"improves":[130],"baselines":[132],"closely":[134],"tracks":[135],"Oracle-Neyman,":[136],"delivering":[137],"up":[138],"28%":[140],"MSE":[141],"reduction":[142],"over":[143],"Uniform":[144],"Sampling":[145],"an":[147],"average":[148],"22.9%":[150],"budget":[151],"savings.":[152]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-13T00:00:00"}