{"id":"W6893445643","doi":"10.5281/zenodo.16882228","title":"Dataset of Kaggle notebooks for reproducibility","year":2025,"lang":"en","type":"dataset","venue":"Zenodo (CERN European Organization for Nuclear Research)","topic":"Scientific Computing and Data Management","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"European Commission","keywords":"Reproducibility; Scripting language; Upload; Limit (mathematics); Open source","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":["metaresearch","open_science"],"consensus_categories":[],"category_scores_codex":[0.002423623,0.002123871,0.001202808,0.005678474,0.001688972,0.002915832,0.003242694,0.002199039,0.0526481],"category_scores_gemma":[0.01359793,0.0007464226,0.001544162,0.007919098,0.0008291439,0.00204844,0.003206376,0.00265677,0.09827974],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001843443,"about_ca_system_score_gemma":0.004144885,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01843879,"about_ca_topic_score_gemma":0.03659331,"domain_scores_codex":[0.9967348,0.0005881892,0.000511219,0.001001072,0.0008112376,0.0003535848],"domain_scores_gemma":[0.9914442,0.001962562,0.0004948713,0.003018929,0.002574577,0.0005049767],"domain_codex":null,"domain_gemma":"reproducibility","domain_candidate":"reproducibility","domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00007914213,0.00003353774,0.0006902143,0.0004036551,0.0000244692,0.00003505777,0.00005611509,0.000305518,0.0002127734,0.000737734,0.9936727,0.003749053],"study_design_scores_gemma":[0.0001651032,0.00001499324,0.003073956,0.0002563531,0.00001951349,0.00007584993,0.0001371267,0.0004261701,0.0007010227,0.002023149,0.9930574,0.00004947232],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.0004088626,0.00007402261,0.0004106863,0.00009710732,0.00005678797,0.00005379987,0.9952809,0.002445332,0.001172501],"genre_scores_gemma":[0.0006196976,0.00003707064,0.001297267,0.0000558434,0.000007630265,0.000234077,0.9966108,0.000441492,0.00069621],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.9975764,"threshold_uncertainty_score":0.1761254,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1748525167998588,"score_gpt":0.3831685802106325,"score_spread":0.2083160634107737,"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."}}