{"id":"W3023058189","doi":"10.1007/978-3-030-47358-7_37","title":"Attending Knowledge Facts with BERT-like Models in Question-Answering: Disappointing Results and Some Explanations","year":2020,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Topic Modeling","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":false,"ca_institutions":"Université de Montréal","funders":"","keywords":"Sentence; Computer science; Question answering; Natural language processing; Set (abstract data type); Artificial intelligence; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01950283,0.001531567,0.002569178,0.001417356,0.002050162,0.006722209,0.005467071,0.008330337,0.01739231],"category_scores_gemma":[0.09941261,0.00176577,0.003621876,0.002473515,0.003896932,0.03222062,0.005035607,0.010319,0.003897948],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002571699,"about_ca_system_score_gemma":0.001413463,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005724824,"about_ca_topic_score_gemma":0.003554145,"domain_scores_codex":[0.9881555,0.007525502,0.0006011586,0.00159438,0.001618016,0.0005054352],"domain_scores_gemma":[0.822356,0.1598622,0.001827454,0.01090418,0.003250772,0.001799409],"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":[0.001589077,0.001054177,0.00387147,0.001502236,0.000486479,0.0004139909,0.003086801,0.08373757,0.002776511,0.5288641,0.06801757,0.3045999],"study_design_scores_gemma":[0.00009657205,0.0001036745,0.0004994714,0.00008572079,0.0001288779,0.0001896942,0.0003140699,0.384739,0.001288417,0.6037899,0.008706626,0.00005801459],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04070272,0.006551139,0.8818763,0.03932671,0.001015234,0.0002015743,0.0008777183,0.003497002,0.02595163],"genre_scores_gemma":[0.6641878,0.00335389,0.3013738,0.004564719,0.002644975,0.0002764906,0.00204535,0.0008584656,0.02069449],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01950283,"threshold_uncertainty_score":0.1031421,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03398356676494375,"score_gpt":0.2610527462836509,"score_spread":0.2270691795187071,"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."}}