{"id":"W4383560524","doi":"10.54254/2755-2721/4/20230490","title":"PreSoramimiset: Establishing dataset for Chinese misheard lyrics generation","year":2023,"lang":"en","type":"article","venue":"Applied and Computational Engineering","topic":"Music and Audio Processing","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Queen's University","funders":"","keywords":"Lyrics; Computer science; Transformer; Python (programming language); Conversation; Natural language processing; Focus (optics); Artificial intelligence; Annotation; Information retrieval; Linguistics","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":[],"consensus_categories":[],"category_scores_codex":[0.000158415,0.00009474704,0.00008768185,0.00008749185,0.0001341437,0.0002925039,0.0001619674,0.00002927736,0.000001060873],"category_scores_gemma":[0.00002794535,0.00009301606,0.00001506112,0.0003182087,0.000007294223,0.0002916847,0.00009220871,0.0000528461,0.000005197838],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000009364107,"about_ca_system_score_gemma":0.00002402828,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000001358311,"about_ca_topic_score_gemma":1.973318e-7,"domain_scores_codex":[0.9993408,0.000002767452,0.0001233079,0.000236538,0.0001405372,0.0001560557],"domain_scores_gemma":[0.9996502,0.0001451613,0.00003005921,0.00009591922,0.00002709433,0.00005152865],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000001493866,0.00000643451,0.00001732014,0.00006120057,0.000009174978,0.000001347162,0.0002207638,0.937925,0.001553749,0.02711984,0.00805274,0.02503091],"study_design_scores_gemma":[0.0002062001,0.000005628236,0.0008274962,0.000006322029,0.00000228863,0.00000336651,0.000004562248,0.9886072,0.0000691649,0.003411891,0.006741628,0.0001142344],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02377849,0.0000491299,0.9751956,0.000413258,0.0001419751,0.0001179676,0.00005733016,0.0001843259,0.00006186388],"genre_scores_gemma":[0.8165821,0.000005587982,0.181268,0.0004788549,0.0003116385,0.0000584349,0.001259942,0.00001487682,0.00002053742],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.7939277,"threshold_uncertainty_score":0.3793085,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01823662911690549,"score_gpt":0.2383724636321235,"score_spread":0.220135834515218,"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."}}