{"id":"W7126165022","doi":"10.21428/594757db.b52553ec","title":"Taxonomic Reasoning for Rare Arthropods: Combining Dense Image Captioning and RAG for Interpretable Classification","year":2025,"lang":"en","type":"article","venue":"","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Guelph","funders":"","keywords":"Context (archaeology); Biodiversity; Closed captioning; Taxonomic rank; Matching (statistics); Interpretability; Contextual image classification; Convolutional neural network","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":[],"consensus_categories":[],"category_scores_codex":[0.0003898082,0.0001231433,0.0001585463,0.0001296994,0.0004626499,0.0003421523,0.0003280178,0.00005552439,0.000005848594],"category_scores_gemma":[0.0003280233,0.000127883,0.00005411528,0.0001774192,0.00005083371,0.0003893712,0.0001436325,0.0001075626,0.000006747369],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006798773,"about_ca_system_score_gemma":0.000071494,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001190904,"about_ca_topic_score_gemma":0.00001946505,"domain_scores_codex":[0.9989755,0.00003697672,0.0002477829,0.0004647432,0.0000527585,0.0002223012],"domain_scores_gemma":[0.9987887,0.0005105002,0.0001184496,0.0003991419,0.0001331373,0.00005006136],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00003239041,0.00004917337,0.004022855,0.00009916758,0.00004251782,1.880551e-7,0.0005327055,0.0007054518,0.01270816,0.9088528,0.00322637,0.0697282],"study_design_scores_gemma":[0.0006935908,0.0000377034,0.00730025,0.00007440389,0.00001489687,0.000003133159,0.0001932782,0.9808196,0.0007911661,0.004618683,0.005308033,0.0001452214],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02528145,0.00006227492,0.9679142,0.002690132,0.0001663098,0.0008186679,0.000005094708,0.0002566676,0.002805242],"genre_scores_gemma":[0.5622157,0.000003648674,0.4366599,0.0001595896,0.00002065316,0.0004696953,0.00001278473,0.000007139856,0.0004509155],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9801142,"threshold_uncertainty_score":0.5214916,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01547139791379805,"score_gpt":0.3021148857597908,"score_spread":0.2866434878459928,"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."}}