{"id":"W4386256573","doi":"10.3390/stats6030056","title":"Investigating Self-Rationalizing Models for Commonsense Reasoning","year":2023,"lang":"en","type":"article","venue":"Stats","topic":"Topic Modeling","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université du Québec à Montréal","funders":"Alliance de recherche numérique du Canada; Natural Sciences and Engineering Research Council of Canada; Government of Canada","keywords":"Computer science; Generative grammar; Leverage (statistics); Commonsense reasoning; Language model; Transformer; Artificial intelligence; Natural language processing; Natural language understanding; Generative model; Natural language; Representation (politics)","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00615972,0.0007738612,0.0006055295,0.001189138,0.0005176503,0.002465951,0.001628282,0.001408046,0.004106373],"category_scores_gemma":[0.0288867,0.0004254569,0.001395187,0.0007126151,0.001453333,0.005387784,0.002130072,0.003247522,0.001054067],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001697873,"about_ca_system_score_gemma":0.001275555,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002775546,"about_ca_topic_score_gemma":0.005123243,"domain_scores_codex":[0.997916,0.001045675,0.00009960931,0.0005786843,0.0002322843,0.0001276417],"domain_scores_gemma":[0.9757308,0.0204538,0.000665054,0.002149154,0.0006169192,0.0003842102],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0007733279,0.0006400645,0.02050392,0.0008158574,0.0004083656,0.0003790767,0.002095797,0.5277526,0.007506163,0.2320988,0.01419637,0.1928298],"study_design_scores_gemma":[0.00003780429,0.00003258378,0.000422831,0.0000223684,0.00002071336,0.00003881049,0.00007910766,0.9034364,0.001372695,0.09313332,0.00139243,0.00001096333],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.294369,0.001272553,0.6840947,0.004038097,0.0001646635,0.0002582987,0.002454463,0.005307836,0.008040409],"genre_scores_gemma":[0.900293,0.0002469989,0.09315348,0.0004929931,0.0000810124,0.0001600872,0.00340137,0.0003339786,0.001837041],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.00615972,"threshold_uncertainty_score":0.03257614,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09268685221712009,"score_gpt":0.3107939534302409,"score_spread":0.2181071012131208,"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."}}