{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.006442786,0.001440477,0.001650177,0.002841562,0.0005811271,0.001598665,0.001855404,0.00199922,0.003433276],"category_scores_gemma":[0.01529088,0.0004617392,0.001117921,0.001830002,0.0008428487,0.005085574,0.001364645,0.001980552,0.00227001],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001874855,"about_ca_system_score_gemma":0.002016625,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01800929,"about_ca_topic_score_gemma":0.01947739,"domain_scores_codex":[0.9975994,0.0009532472,0.0002397164,0.0004999479,0.0005136487,0.0001940168],"domain_scores_gemma":[0.9915488,0.006217033,0.0003346202,0.0008155747,0.000776412,0.0003074753],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.003321266,0.001936596,0.007569597,0.002389264,0.0008136213,0.0002341748,0.0003636769,0.3495899,0.01613098,0.007196272,0.02018384,0.5902708],"study_design_scores_gemma":[0.0001076411,0.0005708284,0.001233659,0.00002937887,0.00009767923,0.00008022738,0.00007475411,0.9897214,0.005013738,0.001888002,0.001142975,0.00003970241],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.547644,0.01967054,0.3995942,0.001850463,0.0007588896,0.001027467,0.004506966,0.01405698,0.01089049],"genre_scores_gemma":[0.8343191,0.002577289,0.1439724,0.0004756348,0.0003360299,0.0004275146,0.01079225,0.0003251708,0.006774669],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01800929,"threshold_uncertainty_score":0.03580892,"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."}}