{"id":"W3007129127","doi":"10.18653/v1/2020.emnlp-main.713","title":"Unsupervised Question Decomposition for Question Answering","year":2020,"lang":"en","type":"preprint","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"Canadian Institute for Advanced Research","funders":"Samsung Advanced Institute of Technology; Samsung; Open Philanthropy Project; Nvidia; National Science Foundation","keywords":"Question answering; Computer science; Leverage (statistics); Open domain; Artificial intelligence; Heuristic; Baseline (sea); Information retrieval; Machine learning; Natural language processing","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":[0.003150446,0.001407226,0.0008217272,0.002637927,0.0008122284,0.001276668,0.00176211,0.001724827,0.006167814],"category_scores_gemma":[0.01180356,0.0005486594,0.001860868,0.001750341,0.001250391,0.004561599,0.002797516,0.003514785,0.003682173],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00123354,"about_ca_system_score_gemma":0.001328895,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002117397,"about_ca_topic_score_gemma":0.003335868,"domain_scores_codex":[0.9967697,0.001719701,0.0001609391,0.0008672642,0.0003318217,0.0001506184],"domain_scores_gemma":[0.9916398,0.004928872,0.0003793347,0.001812898,0.0009279783,0.0003111289],"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":[0.0003562268,0.0007335073,0.005378826,0.0009858896,0.0001873068,0.0001271432,0.001450329,0.07056061,0.04440766,0.05873626,0.03427282,0.7828035],"study_design_scores_gemma":[0.0000447634,0.0001580849,0.001421421,0.00005581371,0.0000431513,0.0001763212,0.0002310499,0.8413366,0.01442447,0.1262822,0.01579398,0.0000321811],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01729582,0.0005520508,0.9705132,0.0005929556,0.00008421066,0.0002377565,0.0007711309,0.007443211,0.002509791],"genre_scores_gemma":[0.2760113,0.000353275,0.7103137,0.0005773903,0.0001988113,0.0005134593,0.006764779,0.0006522572,0.004614943],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006167814,"threshold_uncertainty_score":0.02063334,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04192274640212695,"score_gpt":0.3217740653162895,"score_spread":0.2798513189141626,"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."}}