{"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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.001955125,0.0004572398,0.0009141639,0.0009397491,0.0004443286,0.0004630767,0.001503406,0.0002055288,0.0002674059],"category_scores_gemma":[0.00005077303,0.0003323409,0.000360825,0.00766397,0.00007539661,0.0009556818,0.001323122,0.0003352796,0.0003317985],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002701835,"about_ca_system_score_gemma":0.0004930939,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0006678725,"about_ca_topic_score_gemma":0.002370242,"domain_scores_codex":[0.9961882,0.0002730199,0.0008059436,0.001321422,0.0003955892,0.001015836],"domain_scores_gemma":[0.9968984,0.0002987045,0.0001477788,0.001994877,0.0003828757,0.0002773708],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0007276023,0.003349179,0.1117892,0.002669551,0.01133069,0.0001461503,0.2730939,0.2369071,0.0004007597,0.162056,0.008944889,0.1885849],"study_design_scores_gemma":[0.0006883912,0.0002114044,0.001123402,0.0005183861,0.0004431194,0.000003894241,0.003947956,0.989237,0.0003674167,0.0004880365,0.002431486,0.0005395114],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09915966,0.001248101,0.8445708,0.0004833023,0.0006590906,0.0005274013,0.00001125783,0.0002174566,0.05312299],"genre_scores_gemma":[0.9475744,0.00001243504,0.0363635,0.0005422087,0.00006680693,0.0000571099,0.000003255862,0.00001396393,0.01536626],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8484148,"threshold_uncertainty_score":0.9999129,"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."}}