{"id":"W4386204668","doi":"10.1145/3603719.3603725","title":"InfoMoD: Information-theoretic Model Diagnostics","year":2023,"lang":"en","type":"article","venue":"","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Automatic summarization; Computer science; Debugging; Data modeling; Data mining; Machine learning; Artificial intelligence; Software engineering; Programming language","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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0003046401,0.00009978049,0.0000870113,0.0002214181,0.0001339142,0.0002761505,0.0008452993,0.00005069507,0.00004459622],"category_scores_gemma":[0.0004513748,0.00009112405,0.00003922759,0.00100819,0.00005168819,0.00204547,0.000357081,0.00008727406,0.008488199],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003008993,"about_ca_system_score_gemma":0.00008726485,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002715451,"about_ca_topic_score_gemma":0.000009262541,"domain_scores_codex":[0.998944,0.00001729677,0.0002699195,0.0001430189,0.0002863248,0.0003393792],"domain_scores_gemma":[0.9989187,0.0002691903,0.00004726875,0.0005249596,0.000141866,0.00009800986],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[6.636516e-7,0.000007429567,0.00005298718,0.00000669141,0.000002352005,0.000004224207,0.001469713,0.04996322,0.00002091223,0.9250576,0.0148441,0.008570031],"study_design_scores_gemma":[0.00002664035,0.00001705146,0.00003786002,0.000004376487,0.000001137756,0.000001683389,0.0001497199,0.803579,0.004843209,0.1882488,0.002981529,0.0001090636],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.006115855,0.000006604097,0.9474574,0.001828419,0.0002148444,0.0001300574,0.000002021616,0.0008573246,0.04338744],"genre_scores_gemma":[0.9611219,0.0001655494,0.03420551,0.00250577,0.00003720629,0.00005324607,0.00001170163,0.000009348308,0.001889731],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9550061,"threshold_uncertainty_score":0.9922838,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02157812953750826,"score_gpt":0.2580000841401901,"score_spread":0.2364219546026818,"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."}}