{"id":"W4416034485","doi":"10.18653/v1/2025.mathnlp-main.9","title":"Formula-Text Cross-Retrieval: A Benchmarking Study of Dense Embedding Methods for Mathematical Information Retrieval","year":2025,"lang":"","type":"article","venue":"","topic":"Mathematics, Computing, and Information Processing","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"CanBiocin (Canada)","funders":"","keywords":"Benchmarking; Embedding; Document retrieval; Relevance (law); Data retrieval; Measure (data warehouse)","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.009597342,0.0006787966,0.001357235,0.001319826,0.001175143,0.002639933,0.001812486,0.0003941227,0.0002489651],"category_scores_gemma":[0.005384081,0.0006418152,0.0004645023,0.002636919,0.0002046132,0.006599974,0.00132683,0.0004783062,0.0000459031],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002655291,"about_ca_system_score_gemma":0.0007275618,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000009480696,"about_ca_topic_score_gemma":0.000001010859,"domain_scores_codex":[0.9918036,0.0003107282,0.005207528,0.0006236901,0.001055316,0.0009991165],"domain_scores_gemma":[0.9906344,0.002812664,0.00257754,0.001243817,0.002504239,0.0002273508],"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":[0.0007890425,0.002100345,0.0003244689,0.01715284,0.0007296229,0.00000164514,0.148343,0.007695492,0.0002472122,0.3063717,0.0009513347,0.5152934],"study_design_scores_gemma":[0.003063365,0.0005780305,0.0001192804,0.00113872,0.000175184,0.00001620119,0.004325584,0.9444703,0.006105705,0.03857293,0.0008747531,0.0005599064],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.05708461,0.000178788,0.9268318,0.0001053901,0.001511642,0.002874071,0.000005106309,0.0002153965,0.01119319],"genre_scores_gemma":[0.3812559,0.00001534554,0.6170385,0.0003729626,0.00009275972,0.00002476404,0.000006378143,0.0000219782,0.001171359],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9367748,"threshold_uncertainty_score":0.9996033,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0317626660135732,"score_gpt":0.4055146190143272,"score_spread":0.373751953000754,"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."}}