{"meta":{"page":1,"per_page":50,"max_per_page":100,"total":4,"total_is_capped":false,"direct_labels_cover":0,"predictions_cover":4,"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","author_layer_release":"2026-06-26"},"query_hash":"709e4a003527","filters":{"venue":"Yearbook of Phraseology"}},"results":[{"id":"W2610219444","doi":"10.1515/phras-2015-0005","title":"Clichés, an Understudied Subclass of Phrasemes","year":2015,"lang":"en","type":"article","venue":"Yearbook of Phraseology","topic":"Lexicography and Language Studies","field":"Arts and Humanities","cited_by":54,"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":"Phrase; Cliché; Section (typography); Referent; Meaning (existential); Linguistics; Typology; Dimension (graph theory); Subclass; Class (philosophy); Complementizer; History; Computer science; Mathematics; Sociology; Syntax; Pure mathematics; Artificial intelligence; Philosophy; Epistemology","authors":[{"name":"Igor Mel’čuk","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.1430376586620777,"gpt":0.2831488872078733,"spread":0.1401112285457956,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001847069,0.0001117081,0.0002968848,0.0001259569,0.0000553535,0.000006095245,0.0001467368,0.00004606081,0.0005820672],"category_scores_gemma":[0.0000463305,0.00009207409,0.00008488846,0.00003821533,0.001006895,0.00009233489,0.00005864949,0.00007233952,0.00002066718],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000009450829,"about_ca_system_score_gemma":0.00004273401,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0006943081,"about_ca_topic_score_gemma":0.002380373,"domain_scores_codex":[0.9992126,0.00007690524,0.0002474367,0.000149628,0.0001339637,0.0001794216],"domain_scores_gemma":[0.9993738,0.00006794689,0.0001395375,0.00020666,0.0001508029,0.00006128242],"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":[0.000211193,0.0004321887,0.006858062,0.00009054889,0.0004042892,0.00002001999,0.1491859,0.000005088621,0.0006767716,0.8260555,0.01489446,0.001166029],"study_design_scores_gemma":[0.009000548,0.006663862,0.007327576,0.0001890792,0.0007091197,0.00001690251,0.361557,0.0001099208,0.01533804,0.1109963,0.4865549,0.00153674],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.738528,0.002419693,0.00002748095,0.0003304211,0.0005245908,0.0001588231,0.00004445261,0.00006587462,0.2579006],"genre_scores_gemma":[0.998094,0.0000260555,0.00009849518,0.0001731576,0.0003038624,0.000006644745,0.0000112322,0.0000111838,0.001275405],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7150592,"threshold_uncertainty_score":0.6373227,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W3216854874","doi":"10.1515/phras-2021-0004","title":"Morphemic and Syntactic Phrasemes","year":2021,"lang":"en","type":"article","venue":"Yearbook of Phraseology","topic":"Lexicography and Language Studies","field":"Arts and Humanities","cited_by":30,"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":"Morpheme; Linguistics; Lexeme; Prosody; Computer science; Syntactic structure; Syntax; Morphophonology; Artificial intelligence; Natural language processing; Philosophy; Phonology","authors":[{"name":"Igor Mel’čuk","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01413740018423048,"gpt":0.2106917609973028,"spread":0.1965543608130723,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.00004315645,0.00007436717,0.0001594424,0.00004963452,0.00007397167,0.00001201795,0.00004272497,0.00002757426,0.003789374],"category_scores_gemma":[0.00004267801,0.00006532265,0.00004882033,0.00002135226,0.0004486172,0.00005086029,0.00005397222,0.00006534342,0.00002391207],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000003249231,"about_ca_system_score_gemma":0.00002011694,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001121961,"about_ca_topic_score_gemma":0.0004372468,"domain_scores_codex":[0.9995374,0.00003256547,0.0001177562,0.00013268,0.0000551127,0.0001244573],"domain_scores_gemma":[0.9996656,0.00009334998,0.00004622645,0.0001151816,0.00005452977,0.00002513504],"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":[0.00005662557,0.0002103948,0.008172155,0.0002600762,0.0004989189,0.0003052753,0.06838869,6.116385e-7,0.01261152,0.8924795,0.01114416,0.00587201],"study_design_scores_gemma":[0.003161982,0.0006038604,0.009690573,0.0003005678,0.0005963643,0.0002411084,0.0698442,0.00004284945,0.07285995,0.05175906,0.7897133,0.00118621],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8308259,0.008139698,0.000008517737,0.000526552,0.0002701121,0.00005429239,0.00002251968,0.0000394441,0.1601129],"genre_scores_gemma":[0.9954021,0.0001556857,0.000094214,0.0003323769,0.0002106553,0.000005321895,0.000006021141,0.00000707516,0.003786538],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8407205,"threshold_uncertainty_score":0.9971213,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2610765827","doi":"10.1515/phras-2013-0003","title":"In support of multiword unit classifications: Corpus and human rating data validate phraseological classifications of three different multiword unit types","year":2013,"lang":"en","type":"article","venue":"Yearbook of Phraseology","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":12,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false},"ca_institutions":"","funders":"Social Sciences and Humanities Research Council of Canada","keywords":"Computer science; Natural language processing; Linguistics; Artificial intelligence","authors":[{"name":"Georgie Columbus","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.1005506097616091,"gpt":0.3419018279530505,"spread":0.2413512181914414,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0003541503,0.0001743922,0.0004298602,0.0002963324,0.00005481088,0.00001991386,0.001436678,0.0001741158,0.0001252262],"category_scores_gemma":[0.0002989626,0.000146512,0.00003435236,0.0003304363,0.0006061214,0.0003938388,0.0008335796,0.0002641047,0.000007911931],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001514015,"about_ca_system_score_gemma":0.00007488408,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005497838,"about_ca_topic_score_gemma":0.0003792571,"domain_scores_codex":[0.9982531,0.0001753454,0.0006876666,0.0004572172,0.0001895185,0.0002371557],"domain_scores_gemma":[0.9976514,0.0003963503,0.0005213391,0.001153936,0.0002133955,0.00006356055],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"observational","study_design_scores_codex":[0.000031425,0.0008030216,0.1600857,0.0002557902,0.00008321515,0.00001601534,0.0008207846,0.00000990937,0.5613127,0.2578139,0.0009975499,0.01777],"study_design_scores_gemma":[0.003415213,0.001192916,0.3892127,0.0004241844,0.0001635216,0.00006669718,0.0003058586,0.05583028,0.3586789,0.1878851,0.001578317,0.001246315],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9242422,0.0008684937,0.07149257,0.001720492,0.0001066768,0.0008028269,0.00007082003,0.0001700444,0.0005258651],"genre_scores_gemma":[0.8928921,0.00002868659,0.1067552,0.00007642867,0.00001752871,0.00003656518,0.00006729953,0.00001027335,0.0001159802],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2291269,"threshold_uncertainty_score":0.5974587,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W4388575640","doi":"10.1515/phras-2023-0004","title":"Hidden in plain sound: overlooked repetition in <i>Just a Minute</i>","year":2023,"lang":"en","type":"article","venue":"Yearbook of Phraseology","topic":"Language, Discourse, Communication Strategies","field":"Arts and Humanities","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":"","keywords":"Repetition (rhetorical device); Word (group theory); Entertainment; Linguistics; Quarter (Canadian coin); Remainder; Psychology; Computer science; Arithmetic; History; Mathematics; Art; Visual arts; Philosophy","authors":[{"name":"Alison Wray","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.05488430567375836,"gpt":0.294742212249516,"spread":0.2398579065757576,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0003052114,0.0001008012,0.0002222612,0.0003383533,0.00003936332,0.00002334801,0.000233968,0.00006893572,0.001585281],"category_scores_gemma":[0.00005691781,0.0001002729,0.00004397486,0.0001061389,0.0003582315,0.0001381407,0.00009404424,0.0001609606,0.0002976331],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003953344,"about_ca_system_score_gemma":0.00005863945,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001516806,"about_ca_topic_score_gemma":0.02693417,"domain_scores_codex":[0.9990255,0.000140537,0.0003236451,0.0001709767,0.0001149,0.0002244036],"domain_scores_gemma":[0.9992977,0.000188671,0.0001045188,0.000354917,0.00003187302,0.00002229166],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00007877062,0.0001624865,0.005735077,0.00007170784,0.00002544933,0.0001087504,0.06831156,0.0000337079,0.001162917,0.9164954,0.007141724,0.0006724438],"study_design_scores_gemma":[0.01001707,0.0007302107,0.07993023,0.0008070728,0.0001012161,0.00003948187,0.316643,0.001030593,0.002481706,0.4497679,0.1366449,0.001806611],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8504534,0.0002246276,0.000002029721,0.0008735755,0.0001501095,0.0001684226,0.00002597249,0.00007700852,0.1480249],"genre_scores_gemma":[0.9938048,0.00005940654,0.00005840199,0.0001631321,0.0001197935,0.00004695024,0.00007623653,0.00001522491,0.005656026],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.4667275,"threshold_uncertainty_score":0.9993274,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null}]}