{"id":"W4414404854","doi":"10.1109/tse.2025.3612253","title":"MetaSel: A Test Selection Approach for Fine-Tuned DNN Models","year":2025,"lang":"en","type":"article","venue":"IEEE Transactions on Software Engineering","topic":"Software Testing and Debugging Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Huawei Technologies (Canada); University of Ottawa","funders":"Alliance de recherche numérique du Canada; Mitacs; Huawei Technologies","keywords":"Covariate; Software deployment; Model selection; Selection (genetic algorithm); Context (archaeology); Subspace topology; Statistical hypothesis testing; Test data; Probability distribution","routes":{"ca_aff":true,"ca_fund":true,"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":[0.0002476995,0.0002523489,0.0002401826,0.0004643758,0.0002321904,0.0001250708,0.0004508317,0.0001255363,0.000001656288],"category_scores_gemma":[0.0001920995,0.0002682194,0.0001644251,0.0009359812,0.00001448208,0.0003693583,0.000004322128,0.000268647,0.000002433326],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001326353,"about_ca_system_score_gemma":0.00008417072,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001892159,"about_ca_topic_score_gemma":0.000001735382,"domain_scores_codex":[0.9987122,0.00001657939,0.0002622356,0.0004924664,0.0001735912,0.0003429308],"domain_scores_gemma":[0.9982699,0.001037212,0.00004260174,0.0004487142,0.0001285699,0.00007302954],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00001042649,0.0002474293,0.00003708783,0.0001601321,0.00007861808,0.000001423182,0.0001020316,0.9570686,0.0005129894,0.001404097,0.002977658,0.03739951],"study_design_scores_gemma":[0.0002998178,0.0001032345,0.00002010445,0.00008086095,0.00003788537,0.0000128168,0.000001121039,0.9746172,0.02222703,0.001988563,0.0003440328,0.0002673359],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.000154686,0.00005240419,0.9813424,0.00005153431,0.0004207463,0.0003924462,0.00001938345,0.01752032,0.00004607132],"genre_scores_gemma":[0.313001,0.000003270593,0.6862107,0.00007738808,0.0000234762,0.000420343,0.000003121638,0.0000244558,0.0002362232],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.3128463,"threshold_uncertainty_score":0.999977,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01927186097653321,"score_gpt":0.2414419262966003,"score_spread":0.2221700653200671,"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."}}