{"id":"W4391836017","doi":"10.48550/arxiv.2402.09299","title":"Trained Without My Consent: Detecting Code Inclusion In Language Models Trained on Code","year":2024,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; Canadian Institute for Advanced Research","keywords":"Code (set theory); Inclusion (mineral); Computer science; Programming language; Natural language processing; Artificial intelligence; Linguistics; Psychology; Social psychology; Philosophy","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01028234,0.0009323317,0.0007347206,0.001193299,0.0007239977,0.001835526,0.001782434,0.001584606,0.004464254],"category_scores_gemma":[0.06601053,0.0005271716,0.0009071087,0.0008064824,0.00104864,0.002983346,0.00292217,0.002630855,0.004590581],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009078096,"about_ca_system_score_gemma":0.003475426,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006459158,"about_ca_topic_score_gemma":0.01242617,"domain_scores_codex":[0.9916829,0.003116409,0.0005995659,0.002254194,0.001902113,0.0004447975],"domain_scores_gemma":[0.9647907,0.01729831,0.002221587,0.01094067,0.004021051,0.0007276755],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.001695516,0.0006464418,0.09812086,0.0007486551,0.0003386374,0.0018021,0.001942862,0.03964058,0.01989106,0.009809454,0.1224881,0.7028758],"study_design_scores_gemma":[0.0002261358,0.0004078649,0.01946077,0.00040576,0.000123543,0.0009637347,0.0006417856,0.8560562,0.02825541,0.03532188,0.05798505,0.0001518802],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3044649,0.001788905,0.5850523,0.007140842,0.001006264,0.001421966,0.03115627,0.05133669,0.01663186],"genre_scores_gemma":[0.7059057,0.0003109153,0.2309639,0.002495975,0.0001934188,0.001210531,0.04891935,0.001563333,0.008436941],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01028234,"threshold_uncertainty_score":0.05437887,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07116136631781075,"score_gpt":0.2415754576448641,"score_spread":0.1704140913270533,"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."}}