{"id":"W7092191125","doi":"10.1109/ms.2025.3621625","title":"What Inputs Drive Effective Large Language Model-Based Unit Test Generation?","year":2025,"lang":"","type":"article","venue":"IEEE Software","topic":"Software Testing and Debugging Techniques","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Unit testing; Correctness; Software quality; Code coverage; Code (set theory); Test (biology); Test Management Approach; Test case","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.005672667,0.001146694,0.0009301038,0.00130683,0.0005600365,0.003236014,0.001721288,0.001336121,0.002606758],"category_scores_gemma":[0.09016071,0.0007571306,0.0005885959,0.00101147,0.001158555,0.004764884,0.001945623,0.001529456,0.000962103],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00140253,"about_ca_system_score_gemma":0.001533916,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001916305,"about_ca_topic_score_gemma":0.002774842,"domain_scores_codex":[0.9940066,0.003221535,0.0003099158,0.0006694297,0.001407341,0.0003851378],"domain_scores_gemma":[0.9331027,0.05263032,0.00430342,0.004259464,0.004809948,0.0008941324],"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":[0.001679678,0.00214076,0.1261286,0.001640054,0.000221398,0.001289467,0.002214549,0.2287816,0.1263768,0.01843471,0.006346722,0.4847457],"study_design_scores_gemma":[0.0001151993,0.0008430194,0.02176249,0.0001956689,0.000147488,0.0003151342,0.0007141578,0.8972227,0.05724274,0.01880781,0.002571489,0.00006208777],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6786712,0.0008875203,0.3036198,0.001598925,0.00006992648,0.000394186,0.0005101923,0.005830725,0.008417576],"genre_scores_gemma":[0.9405611,0.0001628395,0.05786564,0.0001960927,0.00002086812,0.0001320575,0.0004342835,0.0003469649,0.0002801587],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.005672667,"threshold_uncertainty_score":0.03000027,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01701364261331757,"score_gpt":0.2985044136494039,"score_spread":0.2814907710360864,"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."}}