{"id":"https://openalex.org/W7165165314","doi":"https://doi.org/10.48550/arxiv.2606.19057","title":"Quantifying and Auditing LLM Evaluation via Positive--Unlabeled Learning","display_name":"Quantifying and Auditing LLM Evaluation via Positive--Unlabeled Learning","publication_year":2026,"publication_date":"2026-06-17","ids":{"openalex":"https://openalex.org/W7165165314","doi":"https://doi.org/10.48550/arxiv.2606.19057"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.19057","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.19057","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":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.2606.19057","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5138880292","display_name":"Zilong Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Zilong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5084887779","display_name":"Yi-Ting Hung","orcid":"https://orcid.org/0009-0009-5021-6014"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hung, Yi-Ting","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138957279","display_name":"Lei Ding","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ding, Lei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5134798316","display_name":"Chi-Kuang Yeh","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yeh, Chi-Kuang","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":0.2757999897003174,"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":0.2757999897003174,"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/T13629","display_name":"Text Readability and Simplification","score":0.11940000206232071,"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":0.07959999889135361,"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/robustness","display_name":"Robustness (evolution)","score":0.6643999814987183},{"id":"https://openalex.org/keywords/audit","display_name":"Audit","score":0.6614000201225281},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.6071000099182129},{"id":"https://openalex.org/keywords/embedding","display_name":"Embedding","score":0.5985000133514404},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.5023000240325928},{"id":"https://openalex.org/keywords/false-positive-paradox","display_name":"False positive paradox","score":0.5016000270843506},{"id":"https://openalex.org/keywords/training-set","display_name":"Training set","score":0.4316999912261963}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6915000081062317},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.6643999814987183},{"id":"https://openalex.org/C199521495","wikidata":"https://www.wikidata.org/wiki/Q181487","display_name":"Audit","level":2,"score":0.6614000201225281},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.6071000099182129},{"id":"https://openalex.org/C41608201","wikidata":"https://www.wikidata.org/wiki/Q980509","display_name":"Embedding","level":2,"score":0.5985000133514404},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5203999876976013},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.5023000240325928},{"id":"https://openalex.org/C64869954","wikidata":"https://www.wikidata.org/wiki/Q1859747","display_name":"False positive paradox","level":2,"score":0.5016000270843506},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.48100000619888306},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.4316999912261963},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.387800008058548},{"id":"https://openalex.org/C112930515","wikidata":"https://www.wikidata.org/wiki/Q4389547","display_name":"Risk analysis (engineering)","level":1,"score":0.3061999976634979},{"id":"https://openalex.org/C2777601897","wikidata":"https://www.wikidata.org/wiki/Q3409113","display_name":"Presentation (obstetrics)","level":2,"score":0.301800012588501},{"id":"https://openalex.org/C2776240099","wikidata":"https://www.wikidata.org/wiki/Q327018","display_name":"Interrogation","level":2,"score":0.2906999886035919},{"id":"https://openalex.org/C2989486834","wikidata":"https://www.wikidata.org/wiki/Q3808900","display_name":"True positive rate","level":2,"score":0.2856000065803528},{"id":"https://openalex.org/C133462117","wikidata":"https://www.wikidata.org/wiki/Q4929239","display_name":"Data collection","level":2,"score":0.27559998631477356},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.26460000872612},{"id":"https://openalex.org/C58489278","wikidata":"https://www.wikidata.org/wiki/Q1172284","display_name":"Data set","level":2,"score":0.2549999952316284}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.19057","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.19057","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":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.2606.19057","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.19057","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","score":0.6451255083084106,"display_name":"Quality Education"}],"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],"Models":[2],"(LLMs)":[3],"are":[4,20],"increasingly":[5],"used":[6],"as":[7,59],"judges":[8,103],"for":[9],"scalable":[10,124],"evaluation,":[11],"yet":[12],"such":[13],"LLM--as--a--Judge":[14],"systems":[15],"exhibit":[16],"systematic":[17],"biases":[18],"that":[19],"decoupled":[21],"from":[22],"semantic":[23],"quality,":[24],"most":[25,43],"notably":[26],"verbosity":[27],"bias.":[28],"Meanwhile,":[29],"human":[30,57,111],"supervision":[31,58],"is":[32],"costly":[33],"and":[34,46,64,100,118,125],"typically":[35],"selective,":[36],"yielding":[37],"reliable":[38,85],"positive":[39],"judgments":[40],"but":[41],"leaving":[42],"outputs":[44,89],"unlabelled":[45,88],"potentially":[47],"mixed":[48],"in":[49,90],"quality.":[50],"We":[51],"formulate":[52],"LLM":[53],"evaluation":[54],"under":[55],"selective":[56],"a":[60,66,77,84,91,123],"positive--unlabelled":[61],"learning":[62],"problem":[63],"propose":[65],"geometric":[67],"auditing":[68],"framework":[69],"based":[70],"on":[71],"Partial":[72],"Optimal":[73],"Transport.":[74],"By":[75],"aligning":[76],"small":[78],"set":[79],"of":[80,87],"human--verified":[81],"positives":[82],"with":[83,110],"subset":[86],"fixed":[92],"embedding":[93],"space,":[94],"our":[95],"method":[96],"identifies":[97],"human--consistent":[98],"preferences":[99],"corrects":[101],"biased":[102],"without":[104],"retraining.":[105],"Experiments":[106],"demonstrate":[107],"improved":[108],"alignment":[109],"preferences,":[112],"increased":[113],"robustness":[114],"to":[115,129],"presentation":[116],"biases,":[117],"interpretable":[119],"confidence":[120],"estimates,":[121],"offering":[122],"statistically":[126],"grounded":[127],"alternative":[128],"existing":[130],"LLM--as--a--judge":[131],"pipelines.":[132]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-19T00:00:00"}