{"meta":{"page":1,"per_page":50,"max_per_page":100,"total":9,"total_is_capped":false,"direct_labels_cover":0,"predictions_cover":9,"direct_label_status":"direct model label, unvalidated","prediction_status":"machine_predicted_unvalidated (Codex and Gemma teacher distillation)","score_status":"score_only:v0-immature-baseline (scores rank; they never assert a category)","snapshot":{"source":"OpenAlex, pinned release, all 482 partitions","release":"2026-06-24","frame_built":"2026-07-12"},"query_hash":"c357e742c7f2","filters":{"venue":"Journal of Open Humanities Data"}},"results":[{"id":"W4312706369","doi":"10.5334/johd.88","title":"The CONLIT Dataset of Contemporary Literature","year":2022,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Narrative Theory and Analysis","field":"Arts and Humanities","cited_by":4,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"McGill University","funders":"","keywords":"Elite; Biography; Literature; History; Art","retraction":null,"screen_n_in":null,"score":{"opus":0.2284273115647129,"gpt":0.3408972242918083,"spread":0.1124699127270954,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["sts","scholarly_communication","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.002374055,0.0001089995,0.0003223042,0.0001040522,0.001541512,0.001110447,0.004145603,0.00001004821,0.007172388],"category_scores_gemma":[0.00008204076,0.00006741536,0.00007749601,0.00005713921,0.00046664,0.001972249,0.002057286,0.0003840751,0.00000468085],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002092803,"about_ca_system_score_gemma":0.0001464947,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00009311567,"about_ca_topic_score_gemma":0.0001940522,"domain_scores_codex":[0.9982969,0.0004448695,0.0006705691,0.0001216142,0.0003522327,0.0001138731],"domain_scores_gemma":[0.9977241,0.0003122406,0.0009099499,0.0007849648,0.0002386087,0.00003014434],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0002279186,0.00008471308,0.00001529184,0.00001357859,0.0002746137,0.00004771155,0.01658856,0.000002395803,0.00001635944,0.354021,0.6284654,0.0002424492],"study_design_scores_gemma":[0.0003540007,0.0002171723,0.000004222802,0.00003782617,0.00006060473,0.00003229058,0.0709867,0.0000126975,0.00001371287,0.008195461,0.920004,0.0000813485],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"other","genre_gemma":"empirical","genre_scores_codex":[0.05455007,0.0557031,0.00004043668,0.01013836,0.005366814,0.001391222,0.4041722,0.0000324338,0.4686054],"genre_scores_gemma":[0.9576027,0.0002638094,0.00005940261,0.001331688,0.001095706,0.000008877482,0.01047861,0.00002624253,0.029133],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9030526,"threshold_uncertainty_score":0.9999265,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W4308571509","doi":"10.5334/johd.97","title":"CREMMA Medii Aevi: Literary Manuscript Text Recognition in Latin","year":2023,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Université de Montréal","funders":"","keywords":"Latin Americans; Computer science; Representation (politics); Natural language processing; Artificial intelligence; Segmentation; Medieval Latin; Information retrieval; Linguistics; History; Classics; Political science","retraction":null,"screen_n_in":null,"score":{"opus":0.23807547872461,"gpt":0.3631402314952308,"spread":0.1250647527706208,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["scholarly_communication","open_science"],"consensus_categories":[],"category_scores_codex":[0.001623098,0.0001256937,0.0002581006,0.0005204928,0.000107436,0.001511174,0.006673254,0.00005614461,0.0001283424],"category_scores_gemma":[0.0003150031,0.0001059043,0.00003254237,0.0005205374,0.00004553701,0.007549059,0.00351156,0.0004148528,0.00005919283],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006092148,"about_ca_system_score_gemma":0.000118873,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00009399636,"about_ca_topic_score_gemma":0.0001052445,"domain_scores_codex":[0.9984253,0.0001318948,0.0005718817,0.000251868,0.0003971567,0.0002218648],"domain_scores_gemma":[0.998445,0.0001505288,0.0003954865,0.0007826054,0.0001774468,0.00004893091],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00006957789,0.0001396528,0.0004549309,0.0001556782,0.00004966218,0.002183281,0.007019273,0.000001268959,0.001032408,0.01124598,0.4570965,0.5205517],"study_design_scores_gemma":[0.002246122,0.0006883827,0.00476367,0.004090684,0.00005035517,0.0009240578,0.001847095,0.009224102,0.005362446,0.7734202,0.196238,0.00114495],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.4417606,0.09530307,0.2311466,0.07341314,0.02354151,0.007755398,0.002757518,0.005372938,0.1189492],"genre_scores_gemma":[0.2170743,0.001107055,0.7706659,0.004048653,0.001200692,0.00002446938,0.0006487126,0.00007700981,0.005153174],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.7621742,"threshold_uncertainty_score":0.9995254,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W4312215389","doi":"10.5334/johd.94","title":"The TRANSCOMP Dataset of Literary Translations from 120 Languages and a Parallel Collection of English-language Originals","year":2022,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"McGill University","funders":"","keywords":"Linguistics; Computer science; Word (group theory); Metadata; Literary translation; Natural language processing; Literature; History; Art; World Wide Web; Philosophy","retraction":null,"screen_n_in":null,"score":{"opus":0.05700619531125964,"gpt":0.3531706120189455,"spread":0.2961644167076859,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0009834499,0.00009134463,0.0002428509,0.0001270463,0.0003217241,0.0003718863,0.004125839,0.00001966848,0.00008256155],"category_scores_gemma":[0.0001063407,0.00006684343,0.00003268147,0.0002032514,0.00009805195,0.002081309,0.000981051,0.0002728752,8.67767e-8],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002022096,"about_ca_system_score_gemma":0.0001317631,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0004489318,"about_ca_topic_score_gemma":0.0001479819,"domain_scores_codex":[0.9986591,0.0002186719,0.0005098041,0.0001621051,0.0003508145,0.00009954373],"domain_scores_gemma":[0.9981995,0.0003571958,0.0005433396,0.0007226967,0.0001511835,0.00002610049],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"not_applicable","study_design_scores_codex":[0.001535719,0.001192455,0.001178418,0.0004791016,0.001288309,0.0004620322,0.4465984,0.00009943378,0.01593212,0.09410277,0.3038163,0.133315],"study_design_scores_gemma":[0.007027684,0.003077257,0.001187815,0.001033169,0.0006511596,0.0005451445,0.0622686,0.01101246,0.02120046,0.08971287,0.8008808,0.001402567],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"review","genre_gemma":"empirical","genre_scores_codex":[0.241362,0.3859849,0.2463008,0.005111989,0.003205418,0.003408234,0.1109564,0.0002950977,0.003375209],"genre_scores_gemma":[0.665126,0.0007841489,0.3312152,0.0002821283,0.0002673559,0.00001996347,0.002039879,0.00002355029,0.000241739],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.4970645,"threshold_uncertainty_score":0.7666903,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W4319597719","doi":"10.5334/johd.95","title":"MultiHATHI: A Complete Collection of Multilingual Prose Fiction in the HathiTrust Digital Library","year":2023,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"McGill University","funders":"","keywords":"Bespoke; Metadata; Computer science; Digital library; Classifier (UML); English language; World Wide Web; Information retrieval; Natural language processing; Artificial intelligence; Linguistics; Literature; Art; Poetry","retraction":null,"screen_n_in":null,"score":{"opus":0.1386274117001172,"gpt":0.3461929316426599,"spread":0.2075655199425427,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.0009355603,0.00009479409,0.0001961448,0.0003122716,0.0001407586,0.00155337,0.005260484,0.00004035061,0.00001584746],"category_scores_gemma":[0.0002838131,0.00006512127,0.00003520894,0.0006198023,0.00007215431,0.007569914,0.002062877,0.0003140186,0.000004577105],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002493111,"about_ca_system_score_gemma":0.0001463658,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004781633,"about_ca_topic_score_gemma":0.00002171994,"domain_scores_codex":[0.9987667,0.0001044829,0.0004644383,0.0001698661,0.0003583296,0.0001361551],"domain_scores_gemma":[0.9986027,0.000227359,0.0004401445,0.0006062138,0.0001047726,0.00001875006],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0008370737,0.001578193,0.006042131,0.0006639972,0.0002472283,0.001501937,0.09623351,0.0001176216,0.003862146,0.0338819,0.7028102,0.1522241],"study_design_scores_gemma":[0.01256857,0.005299632,0.03168611,0.006282806,0.0002091466,0.005073335,0.04883227,0.2584504,0.02969742,0.1609772,0.4378755,0.003047566],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8269555,0.01240023,0.09231337,0.01548546,0.005837437,0.008461528,0.004531899,0.00219767,0.03181691],"genre_scores_gemma":[0.92865,0.000098731,0.06989359,0.0002677996,0.0001913635,0.000008718805,0.0001670281,0.000018901,0.0007038974],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2649346,"threshold_uncertainty_score":0.9994831,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W3190643984","doi":"10.5334/johd.42","title":"The Corpus for Idiolectal Research (CIDRE)","year":2021,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Authorship Attribution and Profiling","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true},"ca_institutions":"","funders":"Agence Nationale de la Recherche","keywords":"Paratext; Scripting language; Metadata; License; Computer science; World Wide Web; Open source; Style (visual arts); Literature; Linguistics; Art; Programming language; Software; Philosophy","retraction":null,"screen_n_in":null,"score":{"opus":0.6364291660620014,"gpt":0.5070253019086151,"spread":0.1294038641533863,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["scholarly_communication","open_science"],"consensus_categories":[],"category_scores_codex":[0.00602987,0.0000610319,0.0001449529,0.0000552555,0.00102209,0.002854078,0.007023896,0.00002985166,0.00005150675],"category_scores_gemma":[0.001127001,0.00004200391,0.00004596419,0.0002050661,0.00009743274,0.001758859,0.003619271,0.0003737221,0.00001343188],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005003015,"about_ca_system_score_gemma":0.0006641811,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001720555,"about_ca_topic_score_gemma":0.0001098324,"domain_scores_codex":[0.9984014,0.0003626245,0.0003585747,0.0001755522,0.0004431088,0.0002587294],"domain_scores_gemma":[0.9967007,0.001008727,0.0002039329,0.0009718923,0.001057059,0.00005765899],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00006884465,0.00004797125,0.00007516629,0.00001626505,0.00007142233,0.0001126014,0.0009032951,0.000003570487,0.0003850298,0.7664054,0.2042795,0.02763091],"study_design_scores_gemma":[0.0003854038,0.0001459062,0.0001044278,0.00005091084,0.00000702195,0.0002474539,0.001144051,0.002178077,0.001680967,0.02973528,0.9642414,0.00007907762],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.009849429,0.01502294,0.8970355,0.03802095,0.007333692,0.001276149,0.000461515,0.00005414722,0.03094565],"genre_scores_gemma":[0.734016,0.003578722,0.1930323,0.004324809,0.005703095,0.0000597439,0.0004585026,0.000098982,0.05872786],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.7599619,"threshold_uncertainty_score":0.9983486,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W4389949574","doi":"10.5334/johd.169","title":"3D Modelling of Jerusalem’s Maghrebi Quarter","year":2023,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"3D Surveying and Cultural Heritage","field":"Earth and Planetary Sciences","cited_by":1,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true},"ca_institutions":"","funders":"Université Gustave Eiffel; Università Degli Studi di Modena e Reggio Emila","keywords":"Quarter (Canadian coin); Geology; Archaeology; History","retraction":null,"screen_n_in":null,"score":{"opus":0.3692056332634672,"gpt":0.3160634521328087,"spread":0.05314218113065855,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.001223436,0.00008486599,0.0002371644,0.00008588917,0.0001453022,0.000261595,0.001733224,0.00002880296,0.002983569],"category_scores_gemma":[0.0000299025,0.00006030846,0.00003947813,0.000134652,0.00006462668,0.001569637,0.0001248474,0.0001534969,0.0001664389],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000002366979,"about_ca_system_score_gemma":0.00005201758,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001110352,"about_ca_topic_score_gemma":0.001180032,"domain_scores_codex":[0.9989107,0.00009211506,0.0004156135,0.0001167263,0.0002944289,0.0001704296],"domain_scores_gemma":[0.9991246,0.0001296897,0.0002565567,0.000341276,0.0000965912,0.0000513239],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0006849413,0.0001482797,0.1856468,0.0003953835,0.0005456427,0.0007237314,0.01339118,0.06301608,0.0001212733,0.0007327669,0.664722,0.06987192],"study_design_scores_gemma":[0.002956467,0.001935165,0.2104051,0.00107429,0.0002606335,0.0004958281,0.04866329,0.2637479,0.0001115748,0.005984973,0.4630494,0.001315333],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9474395,0.001303211,0.0003991558,0.0003236839,0.001105788,0.0001906905,0.001977243,0.00002997044,0.04723078],"genre_scores_gemma":[0.994863,0.000321984,0.001714451,0.00008773471,0.0002131379,1.182723e-7,0.0005175807,0.000004456461,0.002277517],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2016726,"threshold_uncertainty_score":0.9979278,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W4416570569","doi":"10.5334/johd.383","title":"Digital Humanities Training and Workshops: Anonymized Survey Results","year":2025,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Digital Humanities and Scholarship","field":"Arts and Humanities","cited_by":0,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":true},"ca_institutions":"","funders":"Social Sciences and Humanities Research Council of Canada; St. Francis Xavier University; University of Ottawa","keywords":"Certificate; Qualitative property; Digital humanities; Survey data collection; Digital media; Computer-assisted web interviewing; Qualitative research","retraction":null,"screen_n_in":null,"score":{"opus":0.3640158202480061,"gpt":0.3469776897929208,"spread":0.01703813045508523,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.002614337,0.0003750886,0.0008443067,0.0004705866,0.001051576,0.01991446,0.003605151,0.00007535776,0.0009120557],"category_scores_gemma":[0.001062972,0.0003201219,0.0001109665,0.0000840565,0.0009660393,0.01195935,0.002288589,0.0005879591,0.00001556019],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007071863,"about_ca_system_score_gemma":0.0003208656,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003246195,"about_ca_topic_score_gemma":0.005232507,"domain_scores_codex":[0.9970393,0.0001883821,0.001418438,0.0004286908,0.0004728282,0.0004523606],"domain_scores_gemma":[0.9964712,0.001091616,0.0007362909,0.0009237064,0.000682722,0.00009442592],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[0.002221395,0.0003930878,0.0004086111,0.0001272154,0.001074384,0.0002190954,0.05436141,0.000002977734,0.000002472169,0.550038,0.3463345,0.04481684],"study_design_scores_gemma":[0.002453831,0.0003162034,0.0009914879,0.0004395553,0.0001051912,0.00005355598,0.06957553,0.00001717896,0.000002551919,0.008608853,0.9170391,0.00039694],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"other","genre_gemma":"empirical","genre_scores_codex":[0.04908289,0.001874358,0.00002294632,0.0003959169,0.001721573,0.0003882478,0.006105221,0.00004853986,0.9403603],"genre_scores_gemma":[0.7840549,0.0002481329,0.0001195207,0.001306116,0.001014166,0.000006843146,0.001336421,0.0000555465,0.2118583],"genre_candidate":"other","genre_consensus":null,"teacher_disagreement_score":0.7349721,"threshold_uncertainty_score":0.9999251,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W4408373139","doi":"10.5334/johd.262","title":"Data from ARTiculating Canadian Identities: The Canada Council Art Bank’s Mycelial Presence and Influence in the Realm of Canadian Fine Art","year":2024,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Visual Culture and Art Theory","field":"Arts and Humanities","cited_by":0,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":true},"ca_institutions":"","funders":"Social Sciences and Humanities Research Council of Canada","keywords":"Realm; Political science; Art; Public administration; Law","retraction":null,"screen_n_in":null,"score":{"opus":0.2285460099227496,"gpt":0.3058638438127861,"spread":0.07731783389003649,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["scholarly_communication","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.002048909,0.0001531111,0.0002523547,0.0001519505,0.0005661553,0.002371528,0.003397843,0.00002818054,0.002010357],"category_scores_gemma":[0.000540118,0.00009065064,0.00002183164,0.00009863751,0.0003561103,0.002853299,0.0005813559,0.0004051114,0.00001859272],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002245676,"about_ca_system_score_gemma":0.004988631,"about_ca_topic_candidate":true,"about_ca_topic_consensus":true,"about_ca_topic_score_codex":0.996195,"about_ca_topic_score_gemma":0.999985,"domain_scores_codex":[0.9981668,0.0001559167,0.0005743397,0.0002274405,0.0005866583,0.0002888174],"domain_scores_gemma":[0.997975,0.0004662169,0.0001796909,0.0009546434,0.0003031248,0.000121322],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0000126425,0.000008757732,0.0001778533,0.00003475139,0.00008253251,0.0002864452,0.02922783,0.00002104495,0.000008264256,0.03397737,0.9360468,0.0001156767],"study_design_scores_gemma":[0.0001243288,0.00003646476,0.001112689,0.0003876414,0.000070425,0.00005438546,0.03049741,0.0002969279,0.000001916639,0.001653445,0.9656416,0.0001227922],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.613803,0.006619694,0.000002227376,0.01436904,0.002660099,0.0009694737,0.03854699,0.00001274186,0.3230167],"genre_scores_gemma":[0.9883953,0.0001632873,0.00001470115,0.00197164,0.0008528403,0.000004016833,0.0005746041,0.00001695506,0.008006654],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3745923,"threshold_uncertainty_score":0.998902,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W4382727197","doi":"10.5334/johd.109","title":"A Mixed Method Twitter Methodology and Anonymous Corpus","year":2023,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Misinformation and Its Impacts","field":"Social Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Brock University","funders":"","keywords":"Vetting; Anonymity; Computer science; Field (mathematics); Work (physics); Process (computing); World Wide Web; Simple (philosophy); Data science; Internet privacy; Information retrieval; Computer security; Engineering; Mathematics","retraction":null,"screen_n_in":null,"score":{"opus":0.7209269913846231,"gpt":0.5227081836891851,"spread":0.198218807695438,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.00778106,0.00006875711,0.0002439497,0.0001971255,0.0003795064,0.0007519362,0.001425114,0.00005055229,0.001020205],"category_scores_gemma":[0.001622724,0.00005590063,0.00002415488,0.0001805578,0.0001551428,0.002334075,0.0007454415,0.0001564701,0.0000473598],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002952325,"about_ca_system_score_gemma":0.0002378324,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0007405187,"about_ca_topic_score_gemma":0.000705039,"domain_scores_codex":[0.9983313,0.0007067525,0.0003629586,0.00008875324,0.0003075978,0.0002026253],"domain_scores_gemma":[0.9984491,0.0006811149,0.0003524121,0.0002678866,0.0001452625,0.0001042044],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0001032951,0.00002446351,0.0001319465,0.00002209836,0.00009834507,0.00006142068,0.1208106,0.000004140831,0.00008980297,0.04638572,0.7851202,0.04714796],"study_design_scores_gemma":[0.000516592,0.00007914054,0.002486311,0.00002784269,0.00003436911,0.00009091612,0.0927248,0.0000880439,0.00004070457,0.005240121,0.8985665,0.0001046563],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"other","genre_gemma":"methods","genre_scores_codex":[0.3881285,0.001234724,0.02048494,0.02820524,0.007135948,0.001432849,0.0006048884,0.0001763782,0.5525965],"genre_scores_gemma":[0.3039359,0.009918678,0.4442974,0.03131739,0.005752069,0.000008179623,0.0005691686,0.0001492136,0.204052],"genre_candidate":"other","genre_consensus":null,"teacher_disagreement_score":0.4238125,"threshold_uncertainty_score":0.999893,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null}]}