{"id":"W4319451811","doi":"10.48550/arxiv.2206.05480","title":"CodeS: Towards Code Model Generalization Under Distribution Shift","year":2022,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Software Engineering Research","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"JST-Mirai Program; Japan Society for the Promotion of Science; Natural Sciences and Engineering Research Council of Canada; Canadian Institute for Advanced Research","keywords":"Computer science; Python (programming language); Abstract syntax tree; Source code; Programming language; Code refactoring; Syntax; Code (set theory); Java; Artificial intelligence; Software","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":[0.004707425,0.002041883,0.00127083,0.002044047,0.0008700312,0.00236535,0.003163966,0.002513976,0.001940548],"category_scores_gemma":[0.02848705,0.0007516655,0.001836592,0.001475394,0.001829859,0.005027157,0.00438459,0.006113781,0.002184116],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002564928,"about_ca_system_score_gemma":0.003241289,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01003454,"about_ca_topic_score_gemma":0.009480737,"domain_scores_codex":[0.9963943,0.0008854599,0.0001747941,0.001495125,0.0007464018,0.0003038168],"domain_scores_gemma":[0.9882151,0.004838178,0.0007373759,0.004083439,0.001723412,0.0004025732],"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":[0.0007636225,0.0004213699,0.02009157,0.0004317087,0.0003138245,0.0002963657,0.0004750266,0.4563662,0.01216667,0.02119827,0.04822532,0.43925],"study_design_scores_gemma":[0.00003136794,0.00007303053,0.0007567271,0.00002504205,0.0000203231,0.00006968615,0.0000553612,0.9634774,0.00409266,0.02847222,0.002910744,0.00001537859],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1791732,0.001485576,0.7825469,0.002932396,0.0003375719,0.0002508129,0.00414738,0.02440665,0.00471957],"genre_scores_gemma":[0.7674876,0.0007005886,0.1997634,0.002137664,0.0002523774,0.000486588,0.02040697,0.002530333,0.006234454],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01003454,"threshold_uncertainty_score":0.02489555,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09117559834616137,"score_gpt":0.2258600677883055,"score_spread":0.1346844694421441,"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."}}