{"id":"W4416799210","doi":"10.1109/snpd65828.2025.11252591","title":"Making the Case for LLM-Generated Automated Program Repair Benchmarks","year":2025,"lang":"","type":"article","venue":"","topic":"Software Testing and Debugging Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Algoma University","funders":"","keywords":"Overfitting; Benchmark (surveying); Benchmarking; Focus (optics); Quality (philosophy); Quality assurance","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","sts","scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.001933047,0.0004732132,0.0004066895,0.0002782335,0.001551132,0.001236345,0.001369087,0.0003002823,0.00004079746],"category_scores_gemma":[0.00169051,0.0003383889,0.0003609954,0.002330499,0.0002452471,0.0002692268,0.0008429853,0.0003679199,0.000006013076],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001416993,"about_ca_system_score_gemma":0.0005720744,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005386882,"about_ca_topic_score_gemma":0.00007480591,"domain_scores_codex":[0.9966185,0.0003065823,0.0008309723,0.001117482,0.0002330153,0.0008934546],"domain_scores_gemma":[0.9954488,0.00165672,0.0002561224,0.001848133,0.000692699,0.00009755568],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00001711689,0.0003283767,0.0002647486,0.000150464,0.0001156907,0.00044963,0.000294935,0.0001049436,0.000004830112,0.009835409,0.7676113,0.2208226],"study_design_scores_gemma":[0.0003360185,0.0005075391,0.0001173868,0.000412374,0.0001011617,0.0008398434,0.00002785573,0.9723616,0.0002192989,0.01157173,0.01310343,0.0004017188],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.003596717,0.001563237,0.5769847,0.00526861,0.002390051,0.004688204,0.00001552884,0.3958492,0.009643762],"genre_scores_gemma":[0.4155512,0.00000691002,0.5807449,0.001622094,0.00008963631,0.0006682895,0.00001016895,0.00002057817,0.001286309],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9722567,"threshold_uncertainty_score":0.9999068,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04917975613988371,"score_gpt":0.3669347614918263,"score_spread":0.3177550053519426,"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."}}