{"id":"W7118051471","doi":"","title":"How Do Agentic AI Systems Address Performance Optimizations? A BERTopic-Based Analysis of Pull Requests","year":2025,"lang":"","type":"article","venue":"ArXiv.org","topic":"Software System Performance and Reliability","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Winnipeg; University of Windsor; University of Manitoba","funders":"","keywords":"Software; Focus (optics); Empirical research; Software system; Software development; Empirical evidence","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.001214434,0.0005850851,0.001416602,0.001423711,0.0005656003,0.000659884,0.001866975,0.0004337498,0.00005504971],"category_scores_gemma":[0.0002922537,0.0005321817,0.0006668579,0.008243658,0.0003712693,0.001381658,0.0004192284,0.0004688792,0.00004073276],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003081314,"about_ca_system_score_gemma":0.0007959653,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002810887,"about_ca_topic_score_gemma":0.00003682706,"domain_scores_codex":[0.995013,0.0004866476,0.001507727,0.00135485,0.0008102328,0.0008275377],"domain_scores_gemma":[0.9943023,0.0003037617,0.0008133897,0.003036378,0.001323698,0.0002204714],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00003450287,0.0002965171,0.8748292,0.001948497,0.001246845,0.000008663528,0.0004577438,0.1182727,0.00007255738,0.000349664,0.0002572348,0.002225837],"study_design_scores_gemma":[0.0006145712,0.0001451275,0.3652969,0.001031688,0.001364205,0.000002200004,0.00007073645,0.6291675,0.001084032,0.000005724215,0.0008231438,0.0003941209],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8259462,0.003597201,0.1642585,0.001518526,0.003082647,0.0009895858,0.00002870053,0.0001712063,0.0004074421],"genre_scores_gemma":[0.9955796,0.0003668765,0.0006137208,0.00032909,0.0001241456,0.0001438294,0.00003802441,0.00002302953,0.002781709],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5108948,"threshold_uncertainty_score":0.9997129,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02569648182076357,"score_gpt":0.2723796907126856,"score_spread":0.246683208891922,"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."}}