{"id":"W6906290182","doi":"10.17026/ss/tgpdjf","title":"ChiSCor: Children's Story Corpus","year":2024,"lang":"en","type":"dataset","venue":"Leiden Repository (Leiden University)","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"University of Toronto; Nederlandse Organisatie voor Wetenschappelijk Onderzoek; Universiteit Leiden","keywords":"Corpus linguistics; Character (mathematics); Fantasy; Set (abstract data type); Computational linguistics; Natural (archaeology)","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","research_integrity","insufficient_payload"],"consensus_categories":["metaepi_narrow","research_integrity"],"category_scores_codex":[0.0005644536,0.00175622,0.001598073,0.00572889,0.001160734,0.00066051,0.004435008,0.001645498,0.0003747657],"category_scores_gemma":[0.0001571301,0.002094308,0.001164411,0.003385188,0.0008483589,0.001081937,0.00226919,0.004321014,0.07655469],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002955772,"about_ca_system_score_gemma":0.001573563,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003390279,"about_ca_topic_score_gemma":0.001032801,"domain_scores_codex":[0.9918702,0.001154862,0.0008289544,0.002858359,0.001867044,0.001420538],"domain_scores_gemma":[0.9935492,0.0003027495,0.001149542,0.003719525,0.0004255824,0.0008534272],"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":[0.0002401063,0.0002332184,0.002125718,0.0003613969,0.00156986,0.01449309,0.00004040265,0.00002902834,0.0005471149,0.0001807672,0.9801146,0.0000646945],"study_design_scores_gemma":[0.0009525548,0.0002111029,0.004372199,0.0008320394,0.003305394,0.001371012,0.00008495839,0.000005071933,0.0003356984,0.00003234388,0.9863763,0.002121263],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.005461939,0.002131193,0.000004143493,0.00008323095,0.005976902,0.001333598,0.9719868,0.001526641,0.0114956],"genre_scores_gemma":[0.0005188193,0.0002440726,0.0000640005,0.00009412148,0.002456368,0.00000910383,0.8714715,0.0004342576,0.1247078],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.1132122,"threshold_uncertainty_score":0.9996506,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.005647610102114967,"score_gpt":0.1895207032048188,"score_spread":0.1838730931027038,"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."}}