{"id":"W3205791561","doi":"10.48550/arxiv.2110.07752","title":"Hindsight: Posterior-guided training of retrievers for improved open-ended generation","year":2021,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Topic Modeling","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Hindsight bias; Computer science; Context (archaeology); Generator (circuit theory); Labrador Retriever; Relevance (law); Information retrieval; Artificial intelligence; Natural language processing; Psychology; Cognitive psychology; Medicine; Power (physics); Geography; Political science","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0004183867,0.0002643915,0.0004842757,0.0001869784,0.0001373347,0.0002952942,0.002379801,0.0003062831,0.00001462339],"category_scores_gemma":[0.00009602787,0.0003271682,0.000229247,0.0003735618,0.00004323451,0.0006982007,0.002635055,0.0003025088,0.000001284997],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002020035,"about_ca_system_score_gemma":0.0006875233,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001778764,"about_ca_topic_score_gemma":0.00007728929,"domain_scores_codex":[0.9979008,0.000112891,0.0003743362,0.001228185,0.00007934599,0.0003044626],"domain_scores_gemma":[0.9976624,0.00007267637,0.0004502937,0.001310541,0.0003870544,0.0001170068],"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":[0.0002455274,0.0004000255,0.0004443318,0.0006843361,0.0009068643,0.0003007861,0.01393532,0.5711926,0.1218376,0.263737,0.001108982,0.02520665],"study_design_scores_gemma":[0.001012234,0.00006348945,0.00005996149,0.00008234974,0.00005473118,0.000004980453,0.0002242406,0.9864579,0.006480206,0.005048228,0.0001519916,0.000359676],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3091666,0.00002293895,0.6887951,0.0001417454,0.0007499511,0.0005217085,0.00001457375,0.00006170116,0.0005256815],"genre_scores_gemma":[0.9440954,0.00002276931,0.05473666,0.000136506,0.00009308007,0.000002590648,0.0000601705,0.00001725818,0.0008356198],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6349288,"threshold_uncertainty_score":0.999918,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.234326412287494,"score_gpt":0.2359855764146311,"score_spread":0.001659164127137125,"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."}}