{"id":"https://openalex.org/W7213588609","doi":"https://doi.org/10.48550/arxiv.2609.20758","title":"Prediction-Powered Smoothing and Validation for Disaggregated AI Evaluation","display_name":"Prediction-Powered Smoothing and Validation for Disaggregated AI Evaluation","publication_year":2026,"publication_date":"2026-09-17","ids":{"openalex":"https://openalex.org/W7213588609","doi":"https://doi.org/10.48550/arxiv.2609.20758"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.20758","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.20758","pdf_url":"https://arxiv.org/pdf/2609.20758","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":null,"license_id":null,"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/2609.20758","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5100547308","display_name":"Sho Kawano","orcid":"https://orcid.org/0009-0008-1841-8305","observed_orcids":["https://orcid.org/0009-0008-1841-8305"]},"institutions":[{"id":"https://openalex.org/I185103710","display_name":"University of California, Santa Cruz","ror":"https://ror.org/03s65by71","country_code":"US","type":"education","lineage":["https://openalex.org/I185103710"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Kawano, Sho","raw_affiliation_strings":["Department of Statistics , University of California , Santa Cruz , CA , USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Statistics , University of California , Santa Cruz , CA , USA","institution_ids":["https://openalex.org/I185103710"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134780333","display_name":"Zehang Richard Li","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I185103710","display_name":"University of California, Santa Cruz","ror":"https://ror.org/03s65by71","country_code":"US","type":"education","lineage":["https://openalex.org/I185103710"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Li, Zehang Richard","raw_affiliation_strings":["Department of Statistics , University of California , Santa Cruz , CA , USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Statistics , University of California , Santa Cruz , CA , USA","institution_ids":["https://openalex.org/I185103710"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5028367042","display_name":"Paul A. Parker","orcid":"https://orcid.org/0000-0002-3206-6122","observed_orcids":["https://orcid.org/0000-0002-3206-6122"]},"institutions":[{"id":"https://openalex.org/I185103710","display_name":"University of California, Santa Cruz","ror":"https://ror.org/03s65by71","country_code":"US","type":"education","lineage":["https://openalex.org/I185103710"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Parker, Paul A.","raw_affiliation_strings":["Department of Statistics , University of California , Santa Cruz , CA , USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Statistics , University of California , Santa Cruz , CA , USA","institution_ids":["https://openalex.org/I185103710"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I185103710"],"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/T13030","display_name":"Survey Sampling and Estimation Techniques","score":0.875,"subfield":{"id":"https://openalex.org/subfields/2613","display_name":"Statistics and Probability"},"field":{"id":"https://openalex.org/fields/26","display_name":"Mathematics"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T13030","display_name":"Survey Sampling and Estimation Techniques","score":0.875,"subfield":{"id":"https://openalex.org/subfields/2613","display_name":"Statistics and Probability"},"field":{"id":"https://openalex.org/fields/26","display_name":"Mathematics"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11798","display_name":"Optimal Experimental Design Methods","score":0.011300000362098217,"subfield":{"id":"https://openalex.org/subfields/1803","display_name":"Management Science and Operations Research"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T13546","display_name":"Census and Population Estimation","score":0.010099999606609344,"subfield":{"id":"https://openalex.org/subfields/2613","display_name":"Statistics and Probability"},"field":{"id":"https://openalex.org/fields/26","display_name":"Mathematics"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/ai-evaluation","display_name":"AI evaluation","score":0.7760000228881836},{"id":"https://openalex.org/keywords/prediction-powered-inference","display_name":"prediction-powered inference","score":0.7799999713897705},{"id":"https://openalex.org/keywords/small-area-estimation","display_name":"small area estimation","score":0.9860000014305115},{"id":"https://openalex.org/keywords/bayesian-smoothing","display_name":"Bayesian smoothing","score":0.6700000166893005},{"id":"https://openalex.org/keywords/interval-estimation","display_name":"interval estimation","score":0.5770000219345093},{"id":"https://openalex.org/keywords/conversation-types","display_name":"conversation types","score":0.5120000243186951}],"concepts":[{"id":"https://openalex.org/C185429906","wikidata":"https://www.wikidata.org/wiki/Q1130160","display_name":"Estimator","level":2,"score":0.6535999774932861},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6003000140190125},{"id":"https://openalex.org/C41426520","wikidata":"https://www.wikidata.org/wiki/Q1192065","display_name":"Point estimation","level":2,"score":0.5401999950408936},{"id":"https://openalex.org/C3770464","wikidata":"https://www.wikidata.org/wiki/Q775963","display_name":"Smoothing","level":2,"score":0.52920001745224},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.5062999725341797},{"id":"https://openalex.org/C2908647359","wikidata":"https://www.wikidata.org/wiki/Q2625603","display_name":"Population","level":2,"score":0.4465000033378601},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.43209999799728394},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.41019999980926514},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.40230000019073486},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.39579999446868896},{"id":"https://openalex.org/C198531522","wikidata":"https://www.wikidata.org/wiki/Q485146","display_name":"Sample (material)","level":2,"score":0.3944000005722046},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.38909998536109924},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.3562999963760376},{"id":"https://openalex.org/C205167067","wikidata":"https://www.wikidata.org/wiki/Q3300636","display_name":"Interval estimation","level":3,"score":0.3508000075817108},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.3346000015735626},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.32760000228881836},{"id":"https://openalex.org/C2780069185","wikidata":"https://www.wikidata.org/wiki/Q7977945","display_name":"Equivalence (formal languages)","level":2,"score":0.3248000144958496},{"id":"https://openalex.org/C160234255","wikidata":"https://www.wikidata.org/wiki/Q812535","display_name":"Bayesian inference","level":3,"score":0.3116999864578247},{"id":"https://openalex.org/C162376815","wikidata":"https://www.wikidata.org/wiki/Q2158281","display_name":"Frequentist inference","level":4,"score":0.30570000410079956},{"id":"https://openalex.org/C103402496","wikidata":"https://www.wikidata.org/wiki/Q1106171","display_name":"Prediction interval","level":2,"score":0.2904999852180481},{"id":"https://openalex.org/C174348530","wikidata":"https://www.wikidata.org/wiki/Q188635","display_name":"Bridging (networking)","level":2,"score":0.2874999940395355},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.2786000072956085},{"id":"https://openalex.org/C2776459999","wikidata":"https://www.wikidata.org/wiki/Q2119376","display_name":"Fidelity","level":2,"score":0.2784000039100647},{"id":"https://openalex.org/C97256817","wikidata":"https://www.wikidata.org/wiki/Q1462316","display_name":"Spurious relationship","level":2,"score":0.2736000120639801},{"id":"https://openalex.org/C102366305","wikidata":"https://www.wikidata.org/wiki/Q1097688","display_name":"Nonparametric statistics","level":2,"score":0.2728999853134155},{"id":"https://openalex.org/C75917345","wikidata":"https://www.wikidata.org/wiki/Q2725298","display_name":"Sampling bias","level":3,"score":0.26809999346733093},{"id":"https://openalex.org/C137668524","wikidata":"https://www.wikidata.org/wiki/Q189813","display_name":"Censoring (clinical trials)","level":2,"score":0.2676999866962433},{"id":"https://openalex.org/C204323151","wikidata":"https://www.wikidata.org/wiki/Q905424","display_name":"Range (aeronautics)","level":2,"score":0.2623000144958496},{"id":"https://openalex.org/C129848803","wikidata":"https://www.wikidata.org/wiki/Q2564360","display_name":"Sample size determination","level":2,"score":0.26010000705718994},{"id":"https://openalex.org/C70437156","wikidata":"https://www.wikidata.org/wiki/Q7228652","display_name":"Pooling","level":2,"score":0.258899986743927},{"id":"https://openalex.org/C44249647","wikidata":"https://www.wikidata.org/wiki/Q208498","display_name":"Confidence interval","level":2,"score":0.25679999589920044}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.20758","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.20758","pdf_url":"https://arxiv.org/pdf/2609.20758","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2609.20758","doi":"https://doi.org/10.48550/arxiv.2609.20758","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.20758","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:2609.20758","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.20758","pdf_url":"https://arxiv.org/pdf/2609.20758","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7213588609.pdf","grobid_xml":"https://content.openalex.org/works/W7213588609.grobid-xml"},"referenced_works_count":22,"referenced_works":["https://openalex.org/W4320003957","https://openalex.org/W4388327685","https://openalex.org/W4392181972","https://openalex.org/W4394972055","https://openalex.org/W4399365078","https://openalex.org/W4399455122","https://openalex.org/W4402670748","https://openalex.org/W4404125036","https://openalex.org/W4404783102","https://openalex.org/W4407806618","https://openalex.org/W4415795999","https://openalex.org/W4416378452","https://openalex.org/W4417115503","https://openalex.org/W4417134733","https://openalex.org/W7126067104","https://openalex.org/W7131221637","https://openalex.org/W7139967264","https://openalex.org/W7151419658","https://openalex.org/W7156673632","https://openalex.org/W7162773354","https://openalex.org/W7197044228","https://openalex.org/W7203655423"],"related_works":["https://openalex.org/W7163048233","https://openalex.org/W4399455122","https://openalex.org/W7128516680","https://openalex.org/W4408771973","https://openalex.org/W7162773354","https://openalex.org/W4415798772","https://openalex.org/W4399794124","https://openalex.org/W7221166697","https://openalex.org/W7161843939","https://openalex.org/W4396881600"],"abstract_inverted_index":{"Evaluating":[0],"an":[1,88,112,189],"AI":[2],"system":[3],"requires":[4],"disaggregated":[5],"assessment,":[6],"as":[7,13,41,186,188],"performance":[8],"varies":[9],"across":[10,117],"domains":[11],"such":[12],"benchmark":[14,144],"task":[15],"types":[16,19],"or":[17],"conversation":[18],"in":[20,170],"deployed":[21,149],"agents.":[22],"Exhaustive":[23],"testing":[24],"is":[25],"expensive,":[26],"so":[27],"evaluation":[28,39],"rests":[29],"on":[30,84,166],"a":[31,42,64,102,118,126,142],"sample":[32,192],"of":[33,52],"labeled":[34],"units.":[35],"We":[36,140],"treat":[37],"the":[38,162,167,179,196],"set":[40],"finite":[43],"population":[44],"and":[45,49,68,81,93,137,148,172,194],"seek":[46],"accurate":[47],"point":[48,171],"interval":[50,173],"estimates":[51,195],"each":[53,107,155],"domain":[54],"mean.":[55],"Direct":[56],"estimators,":[57],"including":[58],"prediction-powered":[59,99,109],"inference":[60],"(PPI),":[61],"use":[62],"only":[63],"domain's":[65,108],"own":[66],"labels":[67,72],"are":[69,73],"imprecise":[70],"where":[71],"few.":[74],"Small":[75],"area":[76],"estimation":[77,92],"addresses":[78],"this":[79],"problem,":[80],"we":[82,97,124],"build":[83],"it":[85],"to":[86,106],"develop":[87],"integrated":[89],"workflow":[90],"for":[91,133],"validation.":[94],"For":[95,122],"estimation,":[96,174],"propose":[98],"smoothing":[100],"(PP-S),":[101],"Bayesian":[103],"model":[104],"fit":[105],"estimate,":[110],"with":[111,145,156,175],"extension":[113],"that":[114],"borrows":[115],"strength":[116],"reporting":[119],"taxonomy":[120],"(PP-TS).":[121],"validation,":[123],"derive":[125],"new,":[127],"approximately":[128],"unbiased":[129],"design-based":[130],"cross-validation":[131],"score":[132,184],"choosing":[134],"among":[135],"direct":[136,168],"smoothed":[138],"estimators.":[139],"study":[141],"curated":[143],"verifiable":[146],"grading":[147],"agent":[150],"traffic":[151],"graded":[152],"by":[153],"humans,":[154],"every":[157],"outcome":[158],"observed.":[159],"In":[160],"both,":[161],"proposed":[163],"estimators":[164,169],"improve":[165],"near-nominal":[176],"coverage.":[177],"At":[178],"same":[180],"sampling":[181],"budget,":[182],"our":[183],"selects":[185],"well":[187],"independent":[190],"validation":[191],"does":[193],"selected":[197],"estimator's":[198],"error":[199],"far":[200],"more":[201],"accurately.":[202]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-09-19T00:00:00"}