{"id":"W4384009618","doi":"10.1109/msr59073.2023.00033","title":"On Codex Prompt Engineering for OCL Generation: An Empirical Study","year":2023,"lang":"en","type":"article","venue":"","topic":"Software Engineering Research","field":"Computer Science","cited_by":33,"is_retracted":false,"has_abstract":true,"ca_institutions":"Polytechnique Montréal","funders":"Science and Engineering Research Council; Canadian Institute for Advanced Research","keywords":"Computer science; Programming language; Object Constraint Language; Task (project management); Unified Modeling Language; Natural language processing; Syntax; Artificial intelligence; Object (grammar); Software engineering; UML tool; Software","routes":{"ca_aff":true,"ca_fund":true,"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":[0.02632071,0.001605675,0.0008755812,0.00409502,0.001595219,0.00354559,0.003107331,0.002741602,0.002680283],"category_scores_gemma":[0.3441609,0.0008842646,0.001094696,0.003496945,0.002467544,0.007084528,0.003422771,0.003820493,0.002015027],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002127024,"about_ca_system_score_gemma":0.0021717,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01144845,"about_ca_topic_score_gemma":0.008830916,"domain_scores_codex":[0.9651813,0.0188398,0.002517151,0.005077215,0.007764514,0.0006200842],"domain_scores_gemma":[0.4998462,0.4357339,0.01712098,0.02078106,0.02453126,0.001986575],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.00445128,0.003373022,0.2806489,0.01093222,0.0008535417,0.003018378,0.03138385,0.02969931,0.01181403,0.004551969,0.06456532,0.5547081],"study_design_scores_gemma":[0.001252757,0.005806031,0.4127794,0.005563908,0.001320616,0.006985889,0.02061058,0.3313724,0.0276072,0.009681881,0.1761867,0.0008326814],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9515589,0.005868117,0.01916895,0.001147709,0.0002707019,0.001027282,0.005445648,0.007567267,0.007945368],"genre_scores_gemma":[0.9314341,0.001540623,0.03750714,0.0009235149,0.0001330699,0.0009965823,0.02157214,0.002455831,0.00343716],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02632071,"threshold_uncertainty_score":0.1391989,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09103536348884478,"score_gpt":0.3692618199990915,"score_spread":0.2782264565102467,"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."}}