{"id":"W3102645206","doi":"10.18653/v1/2020.emnlp-main.506","title":"Factual Error Correction for Abstractive Summarization Models","year":2020,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":124,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; Mila - Quebec Artificial Intelligence Institute","funders":"Fonds de recherche du Québec – Nature et technologies; Natural Sciences and Engineering Research Council of Canada; Compute Canada","keywords":"Automatic summarization; Computer science; Consistency (knowledge bases); Artificial intelligence; Heuristic; Machine learning; Natural language processing","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00004941618,0.000057729,0.0000630988,0.0000204783,0.00005584149,0.00006051933,0.000196315,0.00003595691,0.00001126032],"category_scores_gemma":[0.00004175667,0.00005441507,0.00003075258,0.00008538196,0.000004665632,0.0006626588,0.00004879919,0.00005018571,0.0000146101],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002092495,"about_ca_system_score_gemma":0.00003260279,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002642907,"about_ca_topic_score_gemma":0.000006876999,"domain_scores_codex":[0.9994376,0.000008722721,0.0001105492,0.0002382098,0.00009987396,0.0001049987],"domain_scores_gemma":[0.9996728,0.00005473183,0.00003762801,0.0001158622,0.00006381102,0.0000551559],"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":[0.00003417457,0.00004755824,0.00008119345,0.00002136587,0.00002389667,0.00000131999,0.006875688,0.5940077,0.002160425,0.1958156,0.007801347,0.1931298],"study_design_scores_gemma":[0.0001280858,0.00003926962,0.00004455068,0.000001588194,0.000001871826,6.375917e-7,0.00007927345,0.9917728,0.0027021,0.00455402,0.0006076354,0.00006815512],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.002295746,0.000003397556,0.9904636,0.001931454,0.0004545953,0.0001843301,0.000001283371,0.0002045856,0.004461026],"genre_scores_gemma":[0.9033955,7.402032e-7,0.09531062,0.000798229,0.00008736319,0.00001340805,0.000004294577,0.000004406789,0.0003854415],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9010997,"threshold_uncertainty_score":0.2218982,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08399768128692162,"score_gpt":0.2738264388879877,"score_spread":0.1898287576010661,"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."}}