{"id":"W7116982847","doi":"10.1214/25-aos2550","title":"Semi-supervised U-statistics","year":2025,"lang":"","type":"article","venue":"The Annals of Statistics","topic":"Statistical Methods and Bayesian Inference","field":"Mathematics","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"Science North","funders":"","keywords":"","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":["metaresearch","metaepi_narrow","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.002296369,0.0007065178,0.001438853,0.0002514168,0.0004837153,0.0002002491,0.001353701,0.0003278626,0.002091358],"category_scores_gemma":[0.01564592,0.0005455969,0.0002033582,0.001005152,0.001606772,0.00009615483,0.0004868911,0.0008380163,0.0001389818],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004361912,"about_ca_system_score_gemma":0.0007572951,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001283235,"about_ca_topic_score_gemma":0.00003142541,"domain_scores_codex":[0.994136,0.001032631,0.002248924,0.0006160738,0.000916622,0.001049735],"domain_scores_gemma":[0.9740157,0.02137582,0.0008738667,0.001614565,0.001862857,0.0002571832],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0001466819,0.0003336658,0.00008427911,0.001601235,0.0003705525,0.00002571663,0.0006589036,0.000005329295,0.0001021787,0.6677966,0.1714964,0.1573784],"study_design_scores_gemma":[0.0006398585,0.0003478278,0.001369244,0.0009005059,0.0007465657,0.000005633297,0.0006001882,0.02440293,0.002111762,0.9615633,0.006806867,0.0005053554],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0001901896,0.001162747,0.9606664,0.002887076,0.001026943,0.0007962891,0.01940418,0.00004684164,0.01381934],"genre_scores_gemma":[0.01913381,0.003334795,0.9711072,0.002055714,0.0001714328,0.00003530683,0.00007583702,0.00007889891,0.004007013],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.2937666,"threshold_uncertainty_score":0.9996995,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1395012079858714,"score_gpt":0.4292025418087274,"score_spread":0.289701333822856,"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."}}