{"id":"W4412827352","doi":"10.1145/3744340","title":"AcTracer: Active Testing of Large Language Model via Multi-Stage Sampling","year":2025,"lang":"en","type":"article","venue":"ACM Transactions on Software Engineering and Methodology","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Sampling (signal processing)","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":[0.01174913,0.003559168,0.002495361,0.001351533,0.00099207,0.002277305,0.006859713,0.002505249,0.003747695],"category_scores_gemma":[0.03460121,0.001120668,0.002114166,0.0007569831,0.001614921,0.006085913,0.004505344,0.004445279,0.002830525],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009395629,"about_ca_system_score_gemma":0.002743321,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005552365,"about_ca_topic_score_gemma":0.007547141,"domain_scores_codex":[0.9901515,0.005610242,0.000496982,0.001723137,0.001531195,0.0004870531],"domain_scores_gemma":[0.9675366,0.02500265,0.000778109,0.003425878,0.0024921,0.0007646354],"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":[0.002566691,0.001444163,0.01363012,0.0009079674,0.0008449425,0.0008154339,0.0009951546,0.2393848,0.03441893,0.007808046,0.01990843,0.6772753],"study_design_scores_gemma":[0.0001527707,0.0003432116,0.0004546003,0.00002127103,0.00005514472,0.0001103475,0.00006766679,0.9849727,0.008145188,0.004280479,0.001356515,0.00003996337],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.06420819,0.001347264,0.9052391,0.0005872447,0.0002654319,0.0006600775,0.0006934571,0.02486438,0.002134834],"genre_scores_gemma":[0.5781119,0.0003803353,0.4078541,0.001497251,0.0002118469,0.001693499,0.003951614,0.002447169,0.00385226],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01174913,"threshold_uncertainty_score":0.06213611,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1514059113826057,"score_gpt":0.3741478920576587,"score_spread":0.222741980675053,"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."}}