{"id":"W4400582376","doi":"10.1145/3660807","title":"Do Large Language Models Pay Similar Attention Like Human Programmers When Generating Code?","year":2024,"lang":"en","type":"article","venue":"Proceedings of the ACM on software engineering.","topic":"Software Engineering Research","field":"Computer Science","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Programmer; Interpretability; Computer science; Code (set theory); Programming language; Code generation; Artificial intelligence; Computer security","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":[0.005873315,0.000687982,0.0006284129,0.001012107,0.000537013,0.002671429,0.001231242,0.001536306,0.002746627],"category_scores_gemma":[0.07296127,0.000704008,0.0005386031,0.0006570037,0.001772331,0.005601228,0.001913059,0.001732226,0.001232841],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007546004,"about_ca_system_score_gemma":0.0009158074,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002903264,"about_ca_topic_score_gemma":0.003819678,"domain_scores_codex":[0.9936491,0.002878122,0.0001901709,0.001467203,0.001347884,0.0004676751],"domain_scores_gemma":[0.9576057,0.02762619,0.003808086,0.007168941,0.002585041,0.001205947],"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":[0.001526958,0.0005054955,0.227238,0.001074846,0.0004929402,0.001433561,0.04129446,0.01489269,0.1157985,0.02094825,0.01558417,0.5592101],"study_design_scores_gemma":[0.0004631724,0.001647838,0.2861018,0.0004891637,0.0006633696,0.004270142,0.0231482,0.3503791,0.07297616,0.1463075,0.1130025,0.0005509808],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7016768,0.001054618,0.2643784,0.005417522,0.0002140778,0.0001745179,0.0002560351,0.00656134,0.02026657],"genre_scores_gemma":[0.9665791,0.0002117057,0.02844609,0.001517268,0.00005028462,0.00006760749,0.0002258159,0.0009728273,0.001929218],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.005873315,"threshold_uncertainty_score":0.03106141,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01965023609466952,"score_gpt":0.2678380185332676,"score_spread":0.248187782438598,"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."}}