{"id":"W4415002193","doi":"10.1002/smr.70057","title":"Evaluation and Improvement of Test Selection for Large Language Models","year":2025,"lang":"en","type":"article","venue":"Journal of Software Evolution and Process","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"National Key Research and Development Program of China; Université du Luxembourg","keywords":"Selection (genetic algorithm); Test (biology); Margin (machine learning); Process (computing); Harm; Deep learning; Empirical research; Ground truth","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":[0.000935651,0.0000453247,0.00009401105,0.0001147635,0.00006146672,0.00002615096,0.00008535994,0.00003326361,0.000001018543],"category_scores_gemma":[0.000295142,0.0000391161,0.00002069354,0.0001315129,0.000007048697,0.0003980405,0.00002481016,0.00005406392,2.7279e-8],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000516512,"about_ca_system_score_gemma":0.0001851188,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000005241435,"about_ca_topic_score_gemma":0.000007660397,"domain_scores_codex":[0.9993988,0.00001428904,0.0002200294,0.00009120648,0.0002032958,0.00007237389],"domain_scores_gemma":[0.9989682,0.0000680465,0.0001831065,0.0000497036,0.0007063321,0.00002461413],"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":[0.00009431887,0.0003960605,0.03501006,0.001506076,0.0001389218,7.341876e-7,0.006266804,0.0148018,0.00948125,0.03120513,0.0004628867,0.900636],"study_design_scores_gemma":[0.001011067,0.000206208,0.002216709,0.0001187574,0.00004486185,0.00001058407,0.0002030276,0.9390723,0.00191254,0.05511206,0.00003920159,0.00005275137],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.127159,0.001367348,0.871071,0.0001585605,0.00008706355,0.0001251063,0.00000192946,0.000009840256,0.00002010913],"genre_scores_gemma":[0.9762261,0.00002220116,0.02362805,0.00004894835,0.00003061079,0.000006084463,3.396787e-7,0.000001722401,0.00003593435],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9242705,"threshold_uncertainty_score":0.1595109,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01554874228645765,"score_gpt":0.3038847352079227,"score_spread":0.2883359929214651,"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."}}