{"id":"W2985200526","doi":"10.1075/ijcl.18009.hal","title":"<i>Phonological CorpusTools</i>","year":2019,"lang":"en","type":"article","venue":"International Journal of Corpus Linguistics","topic":"Phonetics and Phonology Research","field":"Psychology","cited_by":15,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"","keywords":"Phonotactics; Phonology; Computer science; Software; Natural language processing; Linguistics; Phonological rule; Artificial intelligence; Neighbourhood (mathematics); Mathematics; Programming language; Philosophy","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["insufficient_payload"],"consensus_categories":["insufficient_payload"],"category_scores_codex":[0.0004052208,0.0001127628,0.0002354115,0.0002141641,0.00002185054,0.00004681556,0.000847523,0.0001485544,0.002417928],"category_scores_gemma":[0.001067121,0.00009576054,0.0001366416,0.00008137782,0.00009686342,0.00001887768,0.0001044735,0.0004821968,0.0009508896],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006242493,"about_ca_system_score_gemma":0.0001137565,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003738144,"about_ca_topic_score_gemma":0.000005277979,"domain_scores_codex":[0.9985339,0.00007933121,0.0005073295,0.0001447478,0.0005019472,0.0002327433],"domain_scores_gemma":[0.9973245,0.0003634794,0.0003806683,0.0001980212,0.001621268,0.0001120515],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"not_applicable","study_design_scores_codex":[0.003271775,0.00179367,0.4406071,0.00002076862,0.001877489,0.005033782,0.003120722,0.000582518,0.01129086,0.4160334,0.09009757,0.02627031],"study_design_scores_gemma":[0.002385205,0.0007253627,0.067315,0.0000473853,0.00003288525,0.001061624,0.000129868,0.0001726557,0.001016246,0.01094916,0.9159503,0.0002143321],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9005131,0.0006822043,0.0009923145,0.0006084213,0.03917392,0.0001053112,0.00003282091,0.00002214549,0.05786972],"genre_scores_gemma":[0.9932379,0.00009643607,0.0006133593,0.0004775957,0.002179011,0.000001335315,0.000005797619,0.00001615407,0.003372468],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8258527,"threshold_uncertainty_score":0.999827,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0293132621904675,"score_gpt":0.3577817425146209,"score_spread":0.3284684803241534,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}