{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0008278,0.0003363309,0.0002747134,0.0003322778,0.0001519387,0.0005372515,0.003229404,0.0001400511,0.00000908375],"category_scores_gemma":[0.002141984,0.0002422394,0.0002193672,0.0008456045,0.00003338783,0.0007291116,0.001515433,0.0006588467,0.0000164669],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001537746,"about_ca_system_score_gemma":0.00004558481,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001007269,"about_ca_topic_score_gemma":6.434004e-7,"domain_scores_codex":[0.997432,0.000008590702,0.0003657264,0.0006663015,0.0008686239,0.0006587916],"domain_scores_gemma":[0.9983423,0.0003008071,0.00008118109,0.000927799,0.000208789,0.0001391019],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00007326598,0.001389173,0.07600979,0.01467724,0.001977617,0.0002105711,0.04138445,0.160623,0.2651814,0.2173081,0.1307551,0.09041028],"study_design_scores_gemma":[0.00116297,0.0005230195,0.00708634,0.003010675,0.00009922574,0.00009254235,0.0002615474,0.9355762,0.02916498,0.00561501,0.01555103,0.001856449],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6104708,0.001684403,0.3773559,0.001750228,0.001875131,0.001156893,0.00004904931,0.005585536,0.00007207042],"genre_scores_gemma":[0.8310743,0.00001166389,0.1677874,0.00009784244,0.0002555148,0.0001374012,0.000005875439,0.0001138837,0.000516105],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7749532,"threshold_uncertainty_score":0.9878235,"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."}}