{"id":"W4385574225","doi":"10.18653/v1/2022.findings-emnlp.31","title":"Lexicon-Enhanced Self-Supervised Training for Multilingual Dense Retrieval","year":2022,"lang":"en","type":"article","venue":"","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Artificial intelligence; Relevance (law); Lexicon; Generator (circuit theory); Training set; Labeled data; Natural language processing; Machine learning; Information retrieval","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006181504,0.0001391669,0.0001623359,0.0001023581,0.0006798501,0.0001064716,0.001129875,0.00003410213,0.000133127],"category_scores_gemma":[0.0001776781,0.0001478768,0.0001025149,0.0004556886,0.00001848099,0.00015351,0.0004321077,0.0002774558,0.00002719623],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00009149394,"about_ca_system_score_gemma":0.0002352554,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006110012,"about_ca_topic_score_gemma":0.000004840242,"domain_scores_codex":[0.9984233,0.00009957678,0.0002515269,0.0005417573,0.0003359431,0.0003478537],"domain_scores_gemma":[0.9987339,0.0004109571,0.00008490237,0.0005837773,0.00008318863,0.0001032754],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0002866748,0.001002691,0.000721003,0.0000801818,0.000187359,0.00001701876,0.0824364,0.04744658,0.1797387,0.2299882,0.001348542,0.4567467],"study_design_scores_gemma":[0.001070868,0.0001616963,0.0006111771,0.000001686991,0.000007363401,0.00001326985,0.0004663989,0.9843068,0.007334538,0.001849867,0.003936794,0.0002395312],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3796904,0.00001498091,0.6154886,0.001534585,0.0002047452,0.0005233831,0.00000741831,0.0008944153,0.001641401],"genre_scores_gemma":[0.6627647,3.572382e-7,0.3361626,0.0005088017,0.00005303653,0.0001601635,0.000008473001,0.0000139878,0.000327891],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9368602,"threshold_uncertainty_score":0.6030241,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03598622992703411,"score_gpt":0.3098080522508624,"score_spread":0.2738218223238283,"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."}}