{"id":"W6967018607","doi":"10.48448/hs80-1a56","title":"GAIA Search: Hugging Face and Pyserini Interoperability for NLP Training Data Exploration","year":2022,"lang":"en","type":"other","venue":"Underline Science Inc.","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Interoperability; Leverage (statistics); Face (sociological concept); Analytics; Scripting language; Operationalization; Domain (mathematical analysis); Scalability; Data curation","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":[0.006926049,0.00158174,0.001271068,0.003500481,0.001532338,0.005627869,0.003539878,0.002276954,0.03362237],"category_scores_gemma":[0.03240179,0.0008794918,0.002307213,0.003341015,0.002024941,0.01087951,0.01157178,0.003644882,0.02428057],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001154167,"about_ca_system_score_gemma":0.00277744,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003383346,"about_ca_topic_score_gemma":0.004477538,"domain_scores_codex":[0.9953413,0.002021699,0.0003530068,0.0008402597,0.001023179,0.0004205857],"domain_scores_gemma":[0.9886664,0.005310567,0.0003250491,0.004348076,0.0007647372,0.0005851554],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.002396313,0.0003265472,0.003607771,0.001798528,0.0002562896,0.0011986,0.008852892,0.007006811,0.01348027,0.08419809,0.4467624,0.4301155],"study_design_scores_gemma":[0.0003509838,0.0002737427,0.005423565,0.0006788971,0.00009503645,0.001187518,0.002872296,0.07665446,0.0222456,0.1920015,0.6978127,0.0004036787],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"methods","genre_gemma":"software","genre_scores_codex":[0.01000047,0.0008664689,0.5389901,0.003295659,0.0006303226,0.0007189694,0.01894714,0.3973587,0.02919216],"genre_scores_gemma":[0.1170563,0.0007709222,0.7327315,0.003017578,0.0003592204,0.003522255,0.05321736,0.06599037,0.02333462],"genre_candidate":"software","genre_consensus":null,"teacher_disagreement_score":0.03362237,"threshold_uncertainty_score":0.112478,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3765175985623099,"score_gpt":0.4075187840282223,"score_spread":0.03100118546591246,"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."}}