{"id":"W6976885201","doi":"10.6084/m9.figshare.5861067.v1","title":"High-definition phonotactic typology in Sahul: Choosing the data-rich approach","year":2018,"lang":"en","type":"article","venue":"Figshare","topic":"Nuclear Engineering Thermal-Hydraulics","field":"Engineering","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Phonotactics; Typology; Linguistic typology; Variation (astronomy); Phonology; Lexical diversity; Rule-based machine translation; Diversity (politics)","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01173365,0.0006432713,0.001070554,0.008442054,0.002257991,0.008041354,0.002919645,0.001805395,0.005832493],"category_scores_gemma":[0.0429914,0.000805677,0.0008719953,0.007660827,0.006033631,0.01186313,0.01039575,0.003083084,0.001611646],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003061776,"about_ca_system_score_gemma":0.002109624,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005771717,"about_ca_topic_score_gemma":0.009298027,"domain_scores_codex":[0.9908766,0.005508073,0.0007480573,0.001503545,0.001054112,0.0003095487],"domain_scores_gemma":[0.9734976,0.01480648,0.001826689,0.006071182,0.002863789,0.0009343102],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"observational","study_design_scores_codex":[0.000383546,0.0002626656,0.09109151,0.002729861,0.0003546955,0.00194318,0.09485898,0.008723401,0.01010161,0.4667386,0.01590179,0.3069102],"study_design_scores_gemma":[0.00004849531,0.0001673427,0.0585902,0.002138267,0.0001097731,0.001260242,0.0669511,0.035193,0.00612757,0.6453274,0.1838315,0.0002550694],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.205652,0.003735141,0.7343761,0.008793257,0.000416876,0.0007651526,0.00703422,0.001304751,0.03792261],"genre_scores_gemma":[0.6204737,0.00118699,0.3653388,0.001320473,0.0001559728,0.001317802,0.006125894,0.0006842217,0.003395997],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01173365,"threshold_uncertainty_score":0.06205422,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06409533963910465,"score_gpt":0.2346857216734122,"score_spread":0.1705903820343075,"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."}}