{"id":"W7082254717","doi":"10.48448/5drz-7e13","title":"DRAMA: Diverse Augmentation from Large Language Models Towards Smaller Generalizable Dense Retrievers","year":2025,"lang":"en","type":"other","venue":"Underline Science Inc.","topic":"Geochemistry and Geologic Mapping","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Inference; Robustness (evolution); Training set; Language model; Bridging (networking); Structured prediction","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00218862,0.002621066,0.001381653,0.0009354643,0.00071409,0.001491201,0.00342099,0.001913848,0.006391796],"category_scores_gemma":[0.006682689,0.001302258,0.00180043,0.000945197,0.001067108,0.004361405,0.004332981,0.004060554,0.006177804],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006770871,"about_ca_system_score_gemma":0.001412595,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005908402,"about_ca_topic_score_gemma":0.01469395,"domain_scores_codex":[0.9989564,0.000323168,0.0000515868,0.0004359884,0.0001371202,0.00009571741],"domain_scores_gemma":[0.9979858,0.0009808356,0.0001050126,0.0006074964,0.0002241165,0.00009663717],"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":[0.0006854101,0.0007404176,0.00387073,0.0006205796,0.0005606125,0.0006053933,0.0005106619,0.2282017,0.03211409,0.01014327,0.04427899,0.6776682],"study_design_scores_gemma":[0.00008519652,0.0001970164,0.0004757535,0.00003442786,0.00006300549,0.0001633615,0.000100405,0.973053,0.00833929,0.01153138,0.005914587,0.00004258516],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05299233,0.001827825,0.9016062,0.0008470835,0.0003521628,0.0002937703,0.002023491,0.03564804,0.004409077],"genre_scores_gemma":[0.465096,0.0008627988,0.4970016,0.002051353,0.0003295315,0.000776037,0.01470615,0.0028117,0.01636491],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006391796,"threshold_uncertainty_score":0.02138269,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02177187000693468,"score_gpt":0.2651422261453536,"score_spread":0.2433703561384189,"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."}}