{"id":"https://openalex.org/W7225628296","doi":"https://doi.org/10.32388/t420iv.9","title":"SAS Multimetric Tribunal: What a Hallucination Benchmark Measures and What It Does Not \u2014 A Format-Controlled Audit of Structural Classification","display_name":"SAS Multimetric Tribunal: What a Hallucination Benchmark Measures and What It Does Not \u2014 A Format-Controlled Audit of Structural Classification","publication_year":2026,"publication_date":"2026-11-09","ids":{"openalex":"https://openalex.org/W7225628296","doi":"https://doi.org/10.32388/t420iv.9"},"language":"en","primary_location":{"id":"doi:10.32388/t420iv.9","doi":"https://doi.org/10.32388/t420iv.9","is_oa":true,"landing_page_url":"https://doi.org/10.32388/t420iv.9","pdf_url":"https://www.qeios.com/read/T420IV.9/pdf","source":{"id":"https://openalex.org/S4210177074","display_name":"Qeios","issn_l":"2632-3834","issn":["2632-3834"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":["anvur-scientific","norway-1"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Qeios","raw_type":"posted-content"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"diamond","oa_url":"https://www.qeios.com/read/T420IV.9/pdf","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5120884568","display_name":"Gonzalo Emir Durante","orcid":"https://orcid.org/0009-0005-0916-3150","observed_orcids":["https://orcid.org/0009-0005-0916-3150"]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Gonzalo Durante","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0009-0005-0916-3150","affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5120884568"],"corresponding_institution_ids":[],"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/T11147","display_name":"Misinformation and Its Impacts","score":0.36039999127388,"subfield":{"id":"https://openalex.org/subfields/3312","display_name":"Sociology and Political Science"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},"topics":[{"id":"https://openalex.org/T11147","display_name":"Misinformation and Its Impacts","score":0.36039999127388,"subfield":{"id":"https://openalex.org/subfields/3312","display_name":"Sociology and Political Science"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.3303000032901764,"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":0.045499999076128006,"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/hallucination-detection","display_name":"hallucination detection","score":0.9620000123977661},{"id":"https://openalex.org/keywords/benchmark-evaluation","display_name":"benchmark evaluation","score":0.33799999952316284},{"id":"https://openalex.org/keywords/halueval","display_name":"HaluEval","score":0.6759999990463257},{"id":"https://openalex.org/keywords/truthfulqa","display_name":"TruthfulQA","score":0.7310000061988831},{"id":"https://openalex.org/keywords/threshold-calibration","display_name":"threshold calibration","score":0.7440000176429749}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6309999823570251},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.5788000226020813},{"id":"https://openalex.org/C42747912","wikidata":"https://www.wikidata.org/wiki/Q1048447","display_name":"Multiplicative function","level":2,"score":0.5062999725341797},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.503000020980835},{"id":"https://openalex.org/C2777530160","wikidata":"https://www.wikidata.org/wiki/Q41796","display_name":"Sentence","level":2,"score":0.3935000002384186},{"id":"https://openalex.org/C2911011789","wikidata":"https://www.wikidata.org/wiki/Q130741","display_name":"Hallucinating","level":2,"score":0.3880000114440918},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3714999854564667},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3682999908924103},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.36800000071525574},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.35850000381469727},{"id":"https://openalex.org/C78469957","wikidata":"https://www.wikidata.org/wiki/Q838062","display_name":"occam","level":2,"score":0.3472000062465668},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.326200008392334},{"id":"https://openalex.org/C79337645","wikidata":"https://www.wikidata.org/wiki/Q779824","display_name":"Outlier","level":2,"score":0.3237999975681305},{"id":"https://openalex.org/C55166926","wikidata":"https://www.wikidata.org/wiki/Q2892946","display_name":"Oracle","level":2,"score":0.31709998846054077},{"id":"https://openalex.org/C2781181686","wikidata":"https://www.wikidata.org/wiki/Q4226068","display_name":"Coherence (philosophical gambling strategy)","level":2,"score":0.30799999833106995},{"id":"https://openalex.org/C118930307","wikidata":"https://www.wikidata.org/wiki/Q600590","display_name":"Tuple","level":2,"score":0.3028999865055084},{"id":"https://openalex.org/C2776836416","wikidata":"https://www.wikidata.org/wiki/Q1364844","display_name":"False alarm","level":2,"score":0.29260000586509705},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.2906000018119812},{"id":"https://openalex.org/C163864269","wikidata":"https://www.wikidata.org/wiki/Q1107106","display_name":"Cohen's kappa","level":2,"score":0.2808000147342682},{"id":"https://openalex.org/C95623464","wikidata":"https://www.wikidata.org/wiki/Q1096149","display_name":"Classifier (UML)","level":2,"score":0.2741999924182892},{"id":"https://openalex.org/C102366305","wikidata":"https://www.wikidata.org/wiki/Q1097688","display_name":"Nonparametric statistics","level":2,"score":0.2680000066757202},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.26350000500679016},{"id":"https://openalex.org/C2776224158","wikidata":"https://www.wikidata.org/wiki/Q187931","display_name":"Phrase","level":2,"score":0.25999999046325684},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.25839999318122864},{"id":"https://openalex.org/C64869954","wikidata":"https://www.wikidata.org/wiki/Q1859747","display_name":"False positive paradox","level":2,"score":0.25290000438690186}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.32388/t420iv.9","doi":"https://doi.org/10.32388/t420iv.9","is_oa":true,"landing_page_url":"https://doi.org/10.32388/t420iv.9","pdf_url":"https://www.qeios.com/read/T420IV.9/pdf","source":{"id":"https://openalex.org/S4210177074","display_name":"Qeios","issn_l":"2632-3834","issn":["2632-3834"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":["anvur-scientific","norway-1"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Qeios","raw_type":"posted-content"}],"best_oa_location":{"id":"doi:10.32388/t420iv.9","doi":"https://doi.org/10.32388/t420iv.9","is_oa":true,"landing_page_url":"https://doi.org/10.32388/t420iv.9","pdf_url":"https://www.qeios.com/read/T420IV.9/pdf","source":{"id":"https://openalex.org/S4210177074","display_name":"Qeios","issn_l":"2632-3834","issn":["2632-3834"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":["anvur-scientific","norway-1"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Qeios","raw_type":"posted-content"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7225628296.pdf","grobid_xml":"https://content.openalex.org/works/W7225628296.grobid-xml"},"referenced_works_count":6,"referenced_works":["https://openalex.org/W3201174429","https://openalex.org/W4309674289","https://openalex.org/W4389518784","https://openalex.org/W4389519254","https://openalex.org/W4401043863","https://openalex.org/W6912122865"],"related_works":["https://openalex.org/W7220746304","https://openalex.org/W7204113320","https://openalex.org/W7203930421","https://openalex.org/W7222108177","https://openalex.org/W7219923487","https://openalex.org/W7155562060","https://openalex.org/W4416037469","https://openalex.org/W7224253124","https://openalex.org/W7166847750","https://openalex.org/W7221667058"],"abstract_inverted_index":{"Hallucination-detection":[0],"systems":[1],"are":[2,338,379,480,564,642,741,767],"usually":[3],"judged":[4],"by":[5,253,397,743,788,959],"a":[6,12,27,38,44,58,76,87,118,294,301,353,398,439,452,470,505,623,717,736,759,1005],"single":[7],"headline":[8],"metric,":[9],"computed":[10],"under":[11,104,234,423],"benchmark":[13,45,90],"design":[14,369],"that":[15,18,127,326,663,676,904,917,973],"determines":[16],"what":[17,43,48],"metric":[19],"can":[20],"show.":[21],"This":[22],"paper\u2019s":[23,967],"contribution":[24,968],"is":[25,34,210,228,251,291,351,428,444,678,731,756,969,1009],"not":[26,345,352,868,889,983],"detection":[28,306,629,677,760,843],"algorithm":[29],"\u2014":[30,36,989,1008],"structural":[31,61,685,801,865,961],"coherence":[32],"scoring":[33],"established":[35],"but":[37],"methodology":[39],"for":[40,315,363],"measuring":[41],"exactly":[42],"proves":[46],"and":[47,72,114,134,152,247,265,290,307,330,350,370,438,499,520,571,603,630,700,729,733,895,907,954],"it":[49,128],"does":[50,344,888,987],"not,":[51],"applied":[52],"to":[53,188,197,243,284,407,488,781,792,816,825,846,876],"the":[54,94,123,156,163,171,179,189,193,203,208,248,271,327,357,364,367,371,486,501,526,657,664,702,710,796,800,826,850,857,871,921,938,941,960,970,991,994,1000,1012],"SAS":[55,863],"Multimetric":[56],"Tribunal:":[57],"thirteen-module,":[59],"GPU-free":[60],"classifier":[62],"combining":[63],"lexical,":[64],"entity-mutation,":[65],"curvature,":[66],"logical-inversion,":[67],"topological-similarity,":[68],"semantic-drift,":[69],"entropy,":[70],"complexity,":[71],"surprisal":[73],"signals":[74,740],"via":[75,545],"cascading":[77],"multiplicative":[78,220,249],"penalty":[79],"(\\(\\kappa_{D}":[80],"=":[81,84,98,287,532,722,805,819,964],"0.56\\),":[82],"\\(\\kappa_{R}":[83],"0.15\\)).":[85],"On":[86],"stratified":[88],"1,800-pair":[89],"(HaluEval-Dialogue,":[91],"HaluEval-QA,":[92],"TruthfulQA),":[93],"tribunal":[95,141,487],"reaches":[96],"F1":[97,244,281,331,601,813],"98.99%":[99],"with":[100,262,304,451,556,606,627,746,799,849,873,878,892,911,932,1011],"zero":[101,308,896],"false":[102,274,608,633,915,935],"positives":[103,275,609],"an":[105,229,447,557],"A\\(\\rightarrow\\)A":[106],"identical-text":[107],"sanity":[108,302],"check.":[109],"We":[110],"report":[111],"this":[112,382,483,645,770],"figure":[113],"then":[115],"show,":[116],"using":[117,431],"format-controlled":[119,624,672,927],"negative":[120,737],"(N1)":[121],"on":[122,237,276,293,310,335,347,409,522,574,656,669,854,926],"same":[124,172,204,1013],"1,800":[125,145],"IDs,":[126],"measures":[129,496],"format":[130],"asymmetry":[131,748],"between":[132,366,698],"source":[133,519],"response":[135,209,506],"rather":[136,577,680],"than":[137,162,578,681],"hallucination":[138,809],"detection.":[139],"The":[140,216,342,385,615,754,773,966],"flags":[142],"1,569":[143,272],"of":[144,219,231,270,359,518,671,705,713,978,997],"verbatim":[146],"correct":[147,157,277],"answers":[148,278,312],"presented":[149],"alone":[150],"(87.2%),":[151],"in":[153,226,356,381,464,482,610,644,769,870,920],"HaluEval-QA":[154],"scores":[155],"answer":[158],"as":[159,300,446,735,758,1004],"_more_":[160],"anomalous":[161],"hallucinated":[164,311],"response.":[165],"Three":[166,639],"further":[167,477,640],"findings":[168],"follow":[169],"from":[170,282,405,508,751,814,844],"audit.":[173],"(i)":[174],"A":[175,256,317,415,581,649,684],"per-module":[176,650,1001],"analysis":[177],"shows":[178,662],"tribunal\u2019s":[180],"seven-module":[181],"core":[182],"has":[183,394],"no":[184,568],"content-sensitive":[185],"module":[186,195,418,450,828],"(prior":[187],"optional":[190,448],"R3.1":[191,386],"guard):":[192],"one":[194],"designed":[196],"capture":[198],"semantic":[199,421],"divergence":[200],"(cre_isi)":[201],"gives":[202],"per-ID":[205],"score":[206],"whether":[207],"byte-identical":[211],"or":[212,883,914],"completely":[213],"different.":[214],"(ii)":[215],"87.97-point":[217],"advantage":[218],"combination":[221],"over":[222,321,584],"weighted":[223],"averaging":[224],"reported":[225,643,734,768,1010],"v4":[227],"artifact":[230,355],"threshold":[232,328,720],"calibration;":[233],"per-scheme":[235],"thresholds":[236],"format-matched":[238],"negatives,":[239],"both":[240,739],"schemes":[241],"converge":[242],"\\(\\approx":[245],"0.69\\)":[246],"scheme":[250],"worse":[252],"AUROC.":[254],"(iii)":[255],"new":[257,416],"differential":[258],"guard":[259,343],"(R3.1,":[260],"grounding-based":[261],"compression":[263,726],"abstention":[264,390],"short-response":[266],"heuristics)":[267],"corrects":[268],"1,133":[269],"format-driven":[273],"alone,":[279],"raising":[280,811],"0.687":[283],"0.833":[285],"(\\(\\Delta\\)F1":[286,818],"\\(+":[288,820],"0.146\\)),":[289],"evaluated":[292,732],"financial":[295],"domain":[296],"(FinQA,":[297],"1,657":[298],"samples)":[299],"check,":[303],"87.0%":[305],"regressions":[309],"(see":[313,361],"\u00a78.9":[314],"framing).":[316],"held-out":[318,336],"50/50":[319],"split":[320],"100":[322],"random":[323],"repetitions":[324],"confirms":[325],"(0.95)":[329],"(0.833":[332],"\\(\\pm\\)":[333],"0.0055":[334],"data)":[337],"stable":[339],"across":[340,587,653],"splits.":[341],"rely":[346],"substring":[348],"matching":[349,789],"constructional":[354],"sense":[358],"\u00a78":[360,362],"distinction":[365],"guard\u2019s":[368,387],"N1":[372,410],"construction).":[373],"V9":[374,473],"ADDENDUM.":[375],"Two":[376,765],"post-peer-review":[377],"advances":[378,461],"documented":[380,481],"revision.":[383,771],"(iv)":[384],"binary":[388],"compression-ratio":[389],"(CR":[391],"\\(>\\)":[392],"0.5)":[393],"been":[395],"replaced":[396],"continuous":[399],"weight":[400],"function":[401],"(\\(\\kappa\\)C),":[402],"reducing":[403],"abstentions":[404],"58":[406],"1":[408],"(\\(-\\)98%)":[411],"without":[412],"regressions.":[413],"(v)":[414],"topological-similarity":[417],"(R3.2)":[419],"detects":[420,864],"equivalence":[422],"paraphrase":[424,699,855],"when":[425],"lexical":[426,665,879,897],"overlap":[427,880,898],"near":[429],"zero,":[430],"canonical":[432,793],"clusters,":[433],"deterministic":[434],"lemmatization,":[435],"passive-voice":[436],"inversion,":[437],"strict":[440],"anchoring":[441],"rule.":[442],"R3.2":[443,774],"integrated":[445],"eighth":[449],"conditional":[453],"gate":[454,778,798],"(PARAPHRASE_RESCUE":[455],"/":[456,458],"CONTRADICTION_VETO":[457],"INACTIVE).":[459],"Both":[460],"were":[462,572],"completed":[463],"approximately":[465],"10":[466],"hours":[467],"total,":[468],"following":[469],"specification-audit-code":[471],"workflow.":[472],"ADDENDUM":[474],"(2026-11-02).":[475],"Four":[476,823],"extension":[478],"modules":[479,563],"revision,":[484],"bringing":[485],"thirteen":[489],"modules.":[490],"(vi)":[491],"R3.3":[492],"Semantic":[493],"Drift":[494],"Trajectory":[495],"ISI":[497],"sentence-by-sentence":[498],"localizes":[500],"break":[502],"point":[503],"where":[504],"diverges":[507],"its":[509],"source.":[510,922],"(vii)":[511],"R3.4":[512],"Entropy":[513],"Differential":[514],"compares":[515],"Shannon":[516],"entropy":[517],"response;":[521],"300":[523],"calibration":[524],"pairs":[525,586,613,928],"control":[527,612],"condition":[528],"yields":[529,595],"\\({|{\\Delta":[530],"H}|}":[531],"0.000\\)":[533],"exactly,":[534],"giving":[535],"perfect":[536],"separation.":[537],"(viii)":[538],"R3.5":[539],"Kolmogorov":[540],"Complexity":[541],"Ratio":[542],"approximates":[543],"complexity":[544],"zlib":[546],"LZ77":[547],"compression.":[548],"(ix)":[549],"R3.6":[550],"Surprisal":[551],"Analysis":[552],"combines":[553],"per-token":[554],"self-information":[555],"out-of-vocabulary":[558],"fraction":[559],"signal.":[560],"All":[561],"four":[562,588,654],"100%":[565],"standard-library,":[566],"add":[567],"external":[569],"dependencies,":[570],"calibrated":[573],"real":[575],"data":[576],"assumed":[579],"thresholds.":[580],"full":[582],"evaluation":[583,626],"30,790":[585],"public":[589],"datasets":[590],"(HaluEval-QA,":[591],"HaluEval-Dialogue,":[592],"TruthfulQA,":[593],"MedHallu)":[594],"aggregate":[596,812],"precision":[597],"0.9989,":[598],"recall":[599],"0.9877,":[600],"0.9933,":[602,605],"accuracy":[604],"22":[607],"20,768":[611],"(0.11%).":[614],"MedHallu":[616,658],"corpus":[617,659],"(10,000":[618,660],"records,":[619],"4-condition":[620],"protocol)":[621],"adds":[622],"medical-domain":[625],"85.10%":[628,925],"0.01%":[631,933],"identity":[632,934],"positives.":[634,936],"V1.4.0":[635],"ANALYSIS":[636],"SUITE":[637],"(2026-11-05).":[638],"analyses":[641],"revision":[646],"(\u00a78.10).":[647],"(x)":[648],"firing":[651],"breakdown":[652],"conditions":[655],"records)":[661],"baseline":[666],"(Jaccard)":[667],"fires":[668],"0.28%":[670],"A\\(\\rightarrow\\)A_\u2032\u2032_":[673,706],"pairs,":[674],"confirming":[675],"content-driven":[679],"length-driven.":[682],"(xi)":[683],"fingerprint":[686,802,851,962],"(12-dimensional":[687],"feature":[688],"vector:":[689],"3":[690],"relational":[691],"+":[692],"9":[693],"absolute)":[694],"achieves":[695,860,948],"zero-overlap":[696],"separation":[697],"hallucination:":[701],"95th":[703],"percentile":[704,712],"(0.2400)":[707],"lies":[708],"below":[709],"5th":[711],"A\\(\\rightarrow\\)B":[714,752,786],"(0.5436),":[715],"yielding":[716],"recommended":[718],"rescue":[719,783,852],"\\(\\kappa_{P}":[721],"0.4039\\).":[723],"(xii)":[724],"Conditional":[725],"(zlib-based":[727],"\\(\\Delta_{SR}\\)":[728],"asymmetry)":[730],"result:":[738],"dominated":[742],"length":[744],"confounds,":[745],"A\\(\\rightarrow\\)A_\u2032_":[747],"(647.9)":[749],"indistinguishable":[750],"(670.6).":[753],"approach":[755],"discarded":[757],"module.":[761],"V1.4.1":[762],"FIX":[763],"(2026-11-06).":[764],"corrections":[766],"(xiii)":[772],"topological":[775],"module\u2019s":[776],"dictionary-based":[777],"was":[779],"found":[780],"accidentally":[782],"hallucinations":[784,866,903,952],"(127/1000":[785],"pairs)":[787],"common":[790],"words":[791],"clusters.":[794],"Replacing":[795],"dictionary":[797],"distance":[803],"(\\(\\kappa_{P}":[804,963],"0.4039\\))":[806],"eliminated":[807],"all":[808],"rescues,":[810],"0.8901":[815],"0.9933":[817],"0.1032\\)).":[821],"(xiv)":[822],"fixes":[824],"source_target_guard":[827],"(numerical":[829],"magnitude":[830],"comparison,":[831],"drug":[832],"name":[833],"gazetteer,":[834],"date":[835],"mutation":[836],"weight,":[837],"causal":[838],"inversion":[839],"detection)":[840],"raised":[841],"adversarial":[842,858,942],"1/5":[845],"5/5.":[847],"Combined":[848],"(5/5":[853],"cases),":[856],"set":[859,943],"10/10.":[861],"SCOPE.":[862],"(content":[867],"present":[869],"source)":[872],"coverage":[874],"limited":[875],"texts":[877],"\\(\\geq\\)":[881],"30%":[882],"shared":[884],"entity":[885],"structure.":[886],"It":[887],"detect:":[890],"paraphrases":[891,957],"total":[893],"restructuring":[894],"(requires":[899],"distributed":[900],"semantics),":[901],"stealth":[902,951],"preserve":[905],"structure":[906],"entities":[908],"(identical":[909],"vocabulary":[910],"mutated":[912],"facts),":[913],"facts":[916],"do":[918,982],"appear":[919],"Declared":[923],"coverage:":[924],"(MedHallu,":[929],"10,000":[930],"records),":[931],"After":[937],"v1.4.1":[939],"fixes,":[940],"(10":[944],"manually":[945],"constructed":[946],"cases)":[947],"10/10:":[949],"5/5":[950,955],"detected":[953],"legitimate":[956],"rescued":[958],"0.4039\\)).":[965],"audit":[971,996],"procedure":[972],"produced":[974],"these":[975],"retirements.":[976],"Five":[977],"v4/v5\u2019s":[979],"quantitative":[980],"claims":[981],"survive":[984,988],"it.":[985],"What":[986],"thresholds,":[990],"\\(\\kappa_{R}\\)":[992],"sweep,":[993],"forensic":[995],"v2":[998],"metrics,":[999],"architecture":[1002],"itself":[1003],"diagnostic":[1006],"instrument":[1007],"standard.":[1014]},"counts_by_year":[],"updated_date":"2026-11-11T12:16:26.678958","created_date":"2026-11-10T00:00:00"}