{"id":"W4385572574","doi":"10.18653/v1/2023.findings-acl.823","title":"Better Language Models of Code through Self-Improvement","year":2023,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Computer science; Automatic summarization; Code (set theory); Benchmark (surveying); Language model; Artificial intelligence; Machine learning; Code generation; Natural language processing; Programming language","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001170404,0.00172109,0.0008582032,0.001142426,0.0004630083,0.001276508,0.001762719,0.001074419,0.002686407],"category_scores_gemma":[0.006959349,0.000633104,0.00140975,0.0008096147,0.0006714768,0.003029018,0.001359359,0.003535656,0.003229583],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001091915,"about_ca_system_score_gemma":0.001815747,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01006988,"about_ca_topic_score_gemma":0.0198475,"domain_scores_codex":[0.9990286,0.0002572662,0.00004238483,0.0004364167,0.0001433227,0.00009206061],"domain_scores_gemma":[0.9974267,0.001417244,0.0001559032,0.0004051333,0.0004939013,0.0001010605],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0004070807,0.0003772501,0.004154528,0.000498234,0.0002142196,0.0002212994,0.0005037209,0.4320919,0.02509852,0.007857306,0.03388092,0.494695],"study_design_scores_gemma":[0.00002129086,0.00003783028,0.0001952276,0.00001325955,0.00001694058,0.00002596511,0.00002404667,0.9909065,0.003260644,0.003470843,0.002016769,0.00001068783],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.08742947,0.002632762,0.8633807,0.001478903,0.0004105162,0.0001914471,0.002564919,0.03619172,0.005719529],"genre_scores_gemma":[0.618499,0.0008460918,0.3542024,0.001157573,0.0002539158,0.0005173928,0.01284454,0.002863227,0.00881592],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01006988,"threshold_uncertainty_score":0.02002251,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03261112974691838,"score_gpt":0.2701301621696314,"score_spread":0.2375190324227131,"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."}}