{"id":"W4412703957","doi":"10.1145/3696630.3728608","title":"A Tool for Generating Exceptional Behavior Tests With Large Language Models","year":2025,"lang":"en","type":"preprint","venue":"","topic":"Software Testing and Debugging Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"Alliance de recherche numérique du Canada; University of Texas at Austin; Cisco Systems; National Science Foundation","keywords":"Computer science; Programming language; Artificial intelligence; Natural language processing","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":[],"consensus_categories":[],"category_scores_codex":[0.0003798487,0.000264179,0.0002571496,0.0001499767,0.0001561544,0.0002931499,0.0009093061,0.0001849839,0.000009552877],"category_scores_gemma":[0.0001266435,0.0002182712,0.000106214,0.0001365338,0.00001840717,0.0001129404,0.001131647,0.0002874399,0.000002386886],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006431923,"about_ca_system_score_gemma":0.0003477796,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00008489726,"about_ca_topic_score_gemma":0.00002739973,"domain_scores_codex":[0.9984275,0.00003290764,0.0002504399,0.0007047799,0.0002656628,0.000318747],"domain_scores_gemma":[0.998418,0.0003040104,0.0001220726,0.0008859762,0.0002221469,0.00004778716],"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":[0.00007380431,0.002284085,0.02575347,0.00205826,0.0003581242,0.0002783556,0.009874085,0.05546411,0.0009507028,0.3088154,0.2164221,0.3776675],"study_design_scores_gemma":[0.0003162087,0.0000848046,0.0004225817,0.0003715039,0.00003700381,0.00002145864,0.000005279979,0.9567052,0.0003383764,0.04110015,0.000083547,0.0005138484],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01448665,0.00008446008,0.9727576,0.0001185809,0.0001665826,0.0007400331,0.00009687284,0.01092919,0.0006199664],"genre_scores_gemma":[0.1171035,0.00000115695,0.879312,0.0004936882,0.0001244683,0.001247599,0.00006034977,0.00001440239,0.001642818],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9012411,"threshold_uncertainty_score":0.8900843,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03533501993292639,"score_gpt":0.3178231966061725,"score_spread":0.2824881766732462,"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."}}