{"id":"W4385573344","doi":"10.18653/v1/2022.findings-emnlp.78","title":"Evaluating Token-Level and Passage-Level Dense Retrieval Models for Math Information Retrieval","year":2022,"lang":"en","type":"article","venue":"","topic":"Handwritten Text Recognition Techniques","field":"Computer Science","cited_by":14,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; Canada First Research Excellence Fund; Compute Canada","keywords":"Computer science; Encoder; Security token; Information retrieval; Domain (mathematical analysis); Human–computer information retrieval; Search engine; Data mining; Artificial intelligence; Mathematics","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":[0.002348119,0.0001781519,0.0002023163,0.0002521288,0.0005634533,0.0003426229,0.000562309,0.0000736857,0.00005969408],"category_scores_gemma":[0.0003273138,0.0001794405,0.00007694145,0.0004863864,0.00003268014,0.002411255,0.0007356498,0.0002306353,0.00001245689],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001488462,"about_ca_system_score_gemma":0.0001763045,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000179298,"about_ca_topic_score_gemma":0.000001756212,"domain_scores_codex":[0.9979162,0.0001359483,0.000499186,0.0003474543,0.0007620888,0.000339083],"domain_scores_gemma":[0.9986357,0.0002829865,0.0002095146,0.0003968173,0.0003670231,0.0001079464],"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":[0.0005622289,0.0001915988,0.0000537072,0.0002072435,0.00008562205,0.00001090198,0.005706125,0.001165014,0.006332105,0.3338294,0.01411013,0.6377459],"study_design_scores_gemma":[0.000990193,0.0005262977,0.0001253626,0.00001353031,0.00001467019,0.00008851293,0.0002205333,0.9103398,0.01166803,0.07417664,0.001520972,0.0003154186],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0311633,0.00003232107,0.9647413,0.0008184343,0.0001763985,0.0008585136,0.0001299807,0.0005353385,0.001544415],"genre_scores_gemma":[0.4285254,0.00001658728,0.5686769,0.001626927,0.00004319802,0.000129146,0.00006092545,0.00001635471,0.0009045327],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9091748,"threshold_uncertainty_score":0.7317369,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1091930906908583,"score_gpt":0.3200185522246476,"score_spread":0.2108254615337893,"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."}}