{"entities":{"Q7370731":{"pageid":31570079,"ns":120,"title":"Item:Q7370731","lastrevid":105464876,"modified":"2026-10-09T18:00:02Z","type":"item","id":"Q7370731","labels":{"en":{"language":"en","value":"evaluatellm"}},"descriptions":{"en":{"language":"en","value":"Statistical Inference for Language Model Evaluations"}},"aliases":{},"claims":{"P31":[{"mainsnak":{"snaktype":"value","property":"P31","hash":"438d0eef0223375d3a92def8ffc6a10a8bf1ca95","datavalue":{"value":{"entity-type":"item","numeric-id":57080,"id":"Q57080"},"type":"wikibase-entityid"},"datatype":"wikibase-item"},"type":"statement","id":"Q7370731$CCE6E938-B849-47E1-9C90-B502ACF8F540","rank":"normal"}],"P1460":[{"mainsnak":{"snaktype":"value","property":"P1460","hash":"908c3454b3659c4b140ccce33c5aee31081edc8d","datavalue":{"value":{"entity-type":"item","numeric-id":5976450,"id":"Q5976450"},"type":"wikibase-entityid"},"datatype":"wikibase-item"},"type":"statement","id":"Q7370731$BCBC4900-34F1-436E-AF8B-E0FC2802A3C6","rank":"normal"}],"P114":[{"mainsnak":{"snaktype":"value","property":"P114","hash":"46bebcbff353210c0cda7715f07aada2d0e07183","datavalue":{"value":{"entity-type":"item","numeric-id":56774,"id":"Q56774"},"type":"wikibase-entityid"},"datatype":"wikibase-item"},"type":"statement","id":"Q7370731$F44C6279-0CFF-455D-959F-B29960A01365","rank":"normal"}],"P1459":[{"mainsnak":{"snaktype":"value","property":"P1459","hash":"0dfe2a0435a14b3b0206ea5afebd3f6f70565644","datavalue":{"value":"Treats language model evaluations as statistical experiments and    supplies the inference they require. Provides central limit theorem and    cluster-robust standard errors for evaluation scores, paired and unpaired    model comparisons, variance decomposition when several responses are drawn    per question, control-variate variance reduction, multiplicity adjustment    across benchmark suites, and power and minimum detectable effect    calculations for planning evaluations, following Miller (2024)    <doi:10.48550/arXiv.2411.00640>. For evaluations scored by a model judge,    implements agreement statistics against a human gold standard and    prediction-powered inference (Angelopoulos et al. 2023)    <doi:10.1126/science.adi6000> with the power-tuned estimator of    Angelopoulos, Bates and Jordan (2023) <doi:10.48550/arXiv.2311.01453>, so a    small set of human labels debiases a large set of judge scores. Leaderboards    are supported through bootstrap rank intervals and Bradley-Terry ratings    (Bradley and Terry 1952) <doi:10.2307/2334029>. Accepts scores from any    evaluation harness.","type":"string"},"datatype":"string"},"type":"statement","id":"Q7370731$54233AD6-FFF7-4C9A-9100-0123F705B873","rank":"normal"}],"P170":[{"mainsnak":{"snaktype":"value","property":"P170","hash":"13d32b34f559818b22efbdf7fd3f69f88435f94e","datavalue":{"value":{"time":"+2026-09-15T00:00:00Z","timezone":0,"before":0,"after":0,"precision":11,"calendarmodel":"http://www.wikidata.org/entity/Q1985727"},"type":"time"},"datatype":"time"},"type":"statement","id":"Q7370731$6D0AE2C2-CD65-4C0D-A83D-F349123C4DFF","rank":"normal"}],"P132":[{"mainsnak":{"snaktype":"value","property":"P132","hash":"10edd141e119b3c38cc2aa087b8fe5922cd0eea3","datavalue":{"value":"0.1.0","type":"string"},"datatype":"string"},"type":"statement","qualifiers":{"P28":[{"snaktype":"value","property":"P28","hash":"b9612b5388e6d69a98fd76a2085eed3f5429ed33","datavalue":{"value":{"time":"+2026-09-15T00:00:00Z","timezone":0,"before":0,"after":0,"precision":11,"calendarmodel":"http://www.wikidata.org/entity/Q1985727"},"type":"time"},"datatype":"time"}]},"qualifiers-order":["P28"],"id":"Q7370731$3F9B22BC-ABD8-419E-856B-C41CD7803D74","rank":"normal"}],"P16":[{"mainsnak":{"snaktype":"value","property":"P16","hash":"f53e1274fb91a56cdf993e0d31ff90d692cec8a9","datavalue":{"value":{"entity-type":"item","numeric-id":7370729,"id":"Q7370729"},"type":"wikibase-entityid"},"datatype":"wikibase-item"},"type":"statement","id":"Q7370731$63A34E8D-E0F2-4A15-8657-E153909CBF38","rank":"normal"}],"P19":[{"mainsnak":{"snaktype":"value","property":"P19","hash":"ddf9cf0f62d15d590c04cf79630170a450cff8c6","datavalue":{"value":{"entity-type":"item","numeric-id":7370729,"id":"Q7370729"},"type":"wikibase-entityid"},"datatype":"wikibase-item"},"type":"statement","id":"Q7370731$5A2E9FDE-DFEB-4197-A7EC-4377E42088D6","rank":"normal"}],"P163":[{"mainsnak":{"snaktype":"value","property":"P163","hash":"798f7287b894eadb2b35f0ab871accc413b115fd","datavalue":{"value":{"entity-type":"item","numeric-id":56842,"id":"Q56842"},"type":"wikibase-entityid"},"datatype":"wikibase-item"},"type":"statement","id":"Q7370731$538DB157-81BC-433E-843D-852728CAFD88","rank":"normal"},{"mainsnak":{"snaktype":"value","property":"P163","hash":"b26fc7b7fb6977874892e45a4763e1c915d89707","datavalue":{"value":{"entity-type":"item","numeric-id":57086,"id":"Q57086"},"type":"wikibase-entityid"},"datatype":"wikibase-item"},"type":"statement","qualifiers":{"P188":[{"snaktype":"value","property":"P188","hash":"b478bb7ef7be4ada70122b7715a4d9c7baaabf40","datavalue":{"value":"https://cran.r-project.org/web/packages/evaluatellm/LICENSE","type":"string"},"datatype":"url"}]},"qualifiers-order":["P188"],"id":"Q7370731$2A26E267-2658-4963-A303-E51BCAB57C86","rank":"normal"}],"P342":[{"mainsnak":{"snaktype":"value","property":"P342","hash":"a0beaece152e3d4cfe5f01b8a4c02caff7e1e6cc","datavalue":{"value":{"entity-type":"item","numeric-id":56774,"id":"Q56774"},"type":"wikibase-entityid"},"datatype":"wikibase-item"},"type":"statement","qualifiers":{"P132":[{"snaktype":"value","property":"P132","hash":"4893605de428ed066db350fabe1a152c123d9814","datavalue":{"value":"\u2265 4.1.0","type":"string"},"datatype":"string"}]},"qualifiers-order":["P132"],"id":"Q7370731$A93B0972-D3D5-494D-A496-648E809B84CC","rank":"normal"}],"P585":[{"mainsnak":{"snaktype":"value","property":"P585","hash":"9e9de166ea7dd567bb0281c05d85582216a2fa83","datavalue":{"value":{"entity-type":"item","numeric-id":57203,"id":"Q57203"},"type":"wikibase-entityid"},"datatype":"wikibase-item"},"type":"statement","qualifiers":{"P132":[{"snaktype":"value","property":"P132","hash":"91c344ea9128a4f30537b5bb66850514871d3769","datavalue":{"value":"\u2265 3.6.0","type":"string"},"datatype":"string"}]},"qualifiers-order":["P132"],"id":"Q7370731$57E6DEDA-3696-4345-AE85-6A25CACE167C","rank":"normal"},{"mainsnak":{"snaktype":"value","property":"P585","hash":"763ab1af0d78af1e43710d686a298c3714e83ab4","datavalue":{"value":{"entity-type":"item","numeric-id":57371,"id":"Q57371"},"type":"wikibase-entityid"},"datatype":"wikibase-item"},"type":"statement","id":"Q7370731$555383BE-AED5-4E23-8F1C-B882CE45E19E","rank":"normal"},{"mainsnak":{"snaktype":"value","property":"P585","hash":"6999ff979737a4c6588605476533045994851be3","datavalue":{"value":{"entity-type":"item","numeric-id":39400,"id":"Q39400"},"type":"wikibase-entityid"},"datatype":"wikibase-item"},"type":"statement","id":"Q7370731$2F57C1A3-4B13-4C0E-8DEC-FD20A6BB92CB","rank":"normal"},{"mainsnak":{"snaktype":"value","property":"P585","hash":"24c3e3d95d6aceb19323da214a060d1bb0535ab2","datavalue":{"value":{"entity-type":"item","numeric-id":57178,"id":"Q57178"},"type":"wikibase-entityid"},"datatype":"wikibase-item"},"type":"statement","id":"Q7370731$317532AB-46B2-4456-98DE-E9EF227E794A","rank":"normal"},{"mainsnak":{"snaktype":"value","property":"P585","hash":"7aac00570a74583c448c8c2a1fe16ed9aaeff406","datavalue":{"value":{"entity-type":"item","numeric-id":57179,"id":"Q57179"},"type":"wikibase-entityid"},"datatype":"wikibase-item"},"type":"statement","id":"Q7370731$A8F7D843-855D-45B8-8EAC-25830F19279E","rank":"normal"}],"P223":[{"mainsnak":{"snaktype":"value","property":"P223","hash":"2a49887379789305e8cd61f1e3a3355d7fbbb612","datavalue":{"value":{"entity-type":"item","numeric-id":7014243,"id":"Q7014243"},"type":"wikibase-entityid"},"datatype":"wikibase-item"},"type":"statement","id":"Q7370731$BD0EAAE7-373E-4CA8-AD17-94A6312C7845","rank":"normal"},{"mainsnak":{"snaktype":"value","property":"P223","hash":"d9bfc055ea318cc14bd32d4be80bf5753c82e251","datavalue":{"value":{"entity-type":"item","numeric-id":7370730,"id":"Q7370730"},"type":"wikibase-entityid"},"datatype":"wikibase-item"},"type":"statement","id":"Q7370731$3BCE02E8-6981-457D-B910-E2E2B9B1E074","rank":"normal"}],"P229":[{"mainsnak":{"snaktype":"value","property":"P229","hash":"671b4e407e3a2bcbaf8bb7d684585a0968485952","datavalue":{"value":"evaluatellm","type":"string"},"datatype":"external-id"},"type":"statement","id":"Q7370731$55499F0C-8D34-4EBC-95ED-11A26A1A490D","rank":"normal"}]},"sitelinks":{"mardi":{"site":"mardi","title":"Evaluatellm","badges":[],"url":"https://portal.mardi4nfdi.de/wiki/Evaluatellm"}}}}}