{"id":"W4361193744","doi":"10.48550/arxiv.2303.14979","title":"Lexicon-Enhanced Self-Supervised Training for Multilingual Dense Retrieval","year":2023,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Relevance (law); Artificial intelligence; Lexicon; Generator (circuit theory); Training set; Labeled data; Natural language processing; Information retrieval; Machine learning","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001879306,0.001254529,0.001405622,0.001573968,0.0006737913,0.0007144092,0.002415828,0.00122187,0.003303951],"category_scores_gemma":[0.006302414,0.0006239262,0.0009440113,0.001562657,0.0009083205,0.002746833,0.001770844,0.001844338,0.003098919],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007760614,"about_ca_system_score_gemma":0.001486104,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006514319,"about_ca_topic_score_gemma":0.01389703,"domain_scores_codex":[0.9987041,0.0004974217,0.00009105529,0.000388001,0.0002037123,0.0001157338],"domain_scores_gemma":[0.9970174,0.001431809,0.0001584389,0.0006705384,0.0006222166,0.00009959804],"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":[0.0005903518,0.0009289712,0.004337232,0.0004455591,0.0001920366,0.0002797498,0.0003781467,0.1116299,0.02704963,0.00403878,0.02838827,0.8217415],"study_design_scores_gemma":[0.00006711134,0.0001484736,0.0006631144,0.0000172216,0.00003283456,0.0001614254,0.00008495594,0.981419,0.009089881,0.00504508,0.003242364,0.0000286762],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.08174849,0.001316412,0.89404,0.000344082,0.0001161355,0.0003290792,0.001054386,0.01656689,0.004484493],"genre_scores_gemma":[0.6100237,0.000439967,0.36802,0.0008751039,0.0002078415,0.0006387017,0.0104432,0.0009799816,0.008371418],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006514319,"threshold_uncertainty_score":0.0129528,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1457918079473519,"score_gpt":0.2567737515440782,"score_spread":0.1109819435967263,"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."}}