{"id":"W6929338111","doi":"10.48448/b970-en40","title":"End-to-End Training of Neural Retrievers for Open-Domain Question Answering","year":2021,"lang":"en","type":"other","venue":"Underline Science Inc.","topic":"Origins and Evolution of Life","field":"Physics and Astronomy","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Question answering; Task (project management); Artificial neural network; Training set; Salient; Supervised learning; Unsupervised learning; Training (meteorology)","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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0005807955,0.0001781989,0.0003101473,0.0002545553,0.0001571224,0.0001275178,0.0005853987,0.00005675081,0.001713201],"category_scores_gemma":[0.00002793921,0.0001764171,0.00007222072,0.0005531663,0.0002395517,0.0001499651,0.0001811817,0.0001272786,0.00000440591],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005861596,"about_ca_system_score_gemma":0.0005899577,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0008237112,"about_ca_topic_score_gemma":0.00009591087,"domain_scores_codex":[0.9986101,0.00002213905,0.000260835,0.00045146,0.0003108679,0.0003445495],"domain_scores_gemma":[0.9992204,0.00003861664,0.0002142756,0.0002761372,0.0001016987,0.0001489049],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00008867049,0.0004741722,0.005997571,0.0002636053,0.0002393087,0.00000702122,0.002161786,0.00176889,0.01863779,0.6173025,0.1471106,0.205948],"study_design_scores_gemma":[0.001188763,0.0002713187,0.0003950744,0.0005835744,0.00004080001,0.00000211207,0.002948357,0.006793998,0.001541146,0.003254495,0.9822928,0.0006875081],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"other","genre_gemma":"empirical","genre_scores_codex":[0.01016726,0.0002754028,0.2857855,0.00179744,0.004057269,0.002030003,0.001394834,0.0001323506,0.69436],"genre_scores_gemma":[0.7758978,0.000004869109,0.1028119,0.0002127089,0.001762292,0.00005146551,0.0003480256,0.0002006975,0.1187103],"genre_candidate":"other","genre_consensus":null,"teacher_disagreement_score":0.8351822,"threshold_uncertainty_score":0.9991994,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04278417251376176,"score_gpt":0.3366154513469272,"score_spread":0.2938312788331654,"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."}}