{"id":"W4381686872","doi":"10.1162/tacl_a_00564","title":"Questions Are All You Need to Train a Dense Passage Retriever","year":2023,"lang":"en","type":"article","venue":"Transactions of the Association for Computational Linguistics","topic":"Topic Modeling","field":"Computer Science","cited_by":28,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; Mila - Quebec Artificial Intelligence Institute","funders":"DeepMind","keywords":"Computer science; Initialization; Code (set theory); Task (project management); Set (abstract data type); Encoder; Artificial intelligence; Information retrieval; Scheme (mathematics); Training set; Domain (mathematical analysis); Language model; Natural language processing; Machine learning","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":[0.0005935582,0.0001076164,0.0001592849,0.0001783604,0.0002773729,0.00006531456,0.0004944394,0.00008611333,0.00000288714],"category_scores_gemma":[0.002474759,0.0001054488,0.0001751254,0.0008877633,0.00001463214,0.00005545412,0.00002851556,0.0001330732,0.00002291463],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002331363,"about_ca_system_score_gemma":0.0001335114,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002653906,"about_ca_topic_score_gemma":0.00002188654,"domain_scores_codex":[0.9986047,0.0000709677,0.0003802735,0.00023298,0.0004897182,0.0002213543],"domain_scores_gemma":[0.9975826,0.0008448114,0.0002673519,0.0002863517,0.0009495089,0.00006942686],"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":[0.000007530342,0.00005542348,0.000622758,0.00001862903,0.00007963835,5.109342e-7,0.0007630007,0.9242567,0.00003743697,0.07090283,0.00260139,0.000654098],"study_design_scores_gemma":[0.0005781777,0.0000456482,0.01313299,0.00004629962,0.00007954909,0.000001398406,0.0001052658,0.8803675,0.0001478213,0.0824341,0.0228431,0.0002181281],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001841569,0.000005962389,0.9889662,0.00672966,0.001397282,0.0003583001,0.0002177889,0.0002163663,0.0002668287],"genre_scores_gemma":[0.8754216,0.000001953463,0.1216641,0.0003697265,0.0001663535,0.00004064702,0.00002391125,0.00001488026,0.002296746],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8735801,"threshold_uncertainty_score":0.4300076,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03032712766975569,"score_gpt":0.2833717000899555,"score_spread":0.2530445724201998,"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."}}