{"id":"W4416017454","doi":"10.1145/3746252.3761189","title":"Adapting Large Language Models to Log Analysis with Interpretable Domain Knowledge","year":2025,"lang":"","type":"article","venue":"","topic":"Software System Performance and Reliability","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"Huawei Technologies (Canada)","funders":"Fundamental Research Funds for the Central Universities; National Natural Science Foundation of China","keywords":"Domain knowledge; Domain (mathematical analysis); Raw data; Domain adaptation; Security token; Natural language; Subject-matter expert; Production (economics)","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":[0.002804153,0.001835364,0.0006375232,0.002085051,0.0005349439,0.001464568,0.002367181,0.001471339,0.002419062],"category_scores_gemma":[0.01765298,0.0005908706,0.001311003,0.001465977,0.0006666973,0.004388162,0.001897976,0.003410404,0.002369434],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001414029,"about_ca_system_score_gemma":0.001657267,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008339828,"about_ca_topic_score_gemma":0.01794174,"domain_scores_codex":[0.9976791,0.001060785,0.0001903317,0.0007113907,0.0002565157,0.0001019396],"domain_scores_gemma":[0.9869394,0.009749398,0.0004363464,0.0016738,0.001036941,0.0001640757],"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":[0.0008819277,0.001738173,0.02499481,0.001519561,0.0003340584,0.00144272,0.001630056,0.2832927,0.02025144,0.004717667,0.05791828,0.6012786],"study_design_scores_gemma":[0.0000736009,0.0001172856,0.003099454,0.00007579999,0.00006115443,0.0002191246,0.0003729214,0.9616027,0.007893929,0.01080471,0.01562629,0.00005299081],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.264141,0.001962941,0.6464823,0.002942654,0.0004018356,0.001161353,0.02146565,0.05518564,0.006256626],"genre_scores_gemma":[0.6145966,0.0006672448,0.3191896,0.00133877,0.0001710258,0.00113844,0.0574184,0.0009589759,0.004520935],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008339828,"threshold_uncertainty_score":0.01658255,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.009992310365318275,"score_gpt":0.2679572590280425,"score_spread":0.2579649486627242,"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."}}