{"id":"W4221053465","doi":"10.1162/tacl_a_00458","title":"Neuro-symbolic Natural Logic with Introspective Revision for Natural Language Inference","year":2022,"lang":"en","type":"article","venue":"Transactions of the Association for Computational Linguistics","topic":"Topic Modeling","field":"Computer Science","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"Queen's University","funders":"","keywords":"Interpretability; Computer science; Artificial intelligence; Inference; Generalization; Introspection; Spurious relationship; Machine learning; Rule of inference; Natural language; Natural (archaeology); Overfitting; Artificial neural network; Cognitive psychology","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":[],"consensus_categories":[],"category_scores_codex":[0.0003000858,0.0001006974,0.00015177,0.00007863023,0.0005593154,0.0000475597,0.0005100151,0.00002596741,0.000003020078],"category_scores_gemma":[0.001183675,0.00008403201,0.000132589,0.0003180458,0.00001903474,0.00006060085,0.00004162686,0.0002334807,4.587802e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002973982,"about_ca_system_score_gemma":0.0001420389,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001622546,"about_ca_topic_score_gemma":0.000009645544,"domain_scores_codex":[0.9988467,0.0000681696,0.0002587635,0.0002271063,0.000441139,0.0001581985],"domain_scores_gemma":[0.9972755,0.001337115,0.0003400169,0.0001989095,0.0008273078,0.00002113221],"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":[0.00004167916,0.000072508,0.0003259231,0.0000287762,0.00006284696,2.321473e-7,0.0009319613,0.8032641,0.00004377042,0.1932195,0.0001317374,0.001876988],"study_design_scores_gemma":[0.0007117529,0.0001374396,0.001409075,0.00001182427,0.00004429512,0.000002812228,0.00009364087,0.9640448,0.0001061777,0.03169777,0.001610784,0.0001295945],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.002208699,0.00007158885,0.9944319,0.0009521377,0.001514728,0.0005296791,0.0001360505,0.00006601575,0.00008927295],"genre_scores_gemma":[0.890322,7.594388e-7,0.1089362,0.0002249046,0.000111588,0.00005202253,0.0000296727,0.000008863304,0.0003140533],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8881133,"threshold_uncertainty_score":0.4301859,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01236504266629197,"score_gpt":0.2721111208052126,"score_spread":0.2597460781389206,"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."}}