{"id":"W4414266922","doi":"10.14778/3750601.3750685","title":"Beyond Quacking: Deep Integration of Language Models and RAG into DuckDB","year":2025,"lang":"en","type":"article","venue":"Proceedings of the VLDB Endowment","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo; Polytechnique Montréal","funders":"","keywords":"Schema (genetic algorithms); Data integration; Data modeling; Relational database; Context (archaeology); SQL; Language model; Context model; Rapid prototyping","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":[0.005099867,0.0009185917,0.001220908,0.002031583,0.0009506996,0.005592314,0.003857737,0.001095568,0.005835359],"category_scores_gemma":[0.01478145,0.001237425,0.001543664,0.001859374,0.001359791,0.007818766,0.005409986,0.002898247,0.003520445],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001991476,"about_ca_system_score_gemma":0.003131998,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01748516,"about_ca_topic_score_gemma":0.02946141,"domain_scores_codex":[0.9970025,0.0008450365,0.0003512854,0.0006520029,0.000999023,0.0001500625],"domain_scores_gemma":[0.9959798,0.001499411,0.0001286256,0.001706006,0.0005069432,0.000179202],"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":[0.0006104496,0.0004493788,0.007800912,0.001209423,0.0004622569,0.0004731969,0.00153819,0.136838,0.005464096,0.1795463,0.1021015,0.5635063],"study_design_scores_gemma":[0.00008479683,0.00005987041,0.0006459354,0.0001811787,0.00006497191,0.0001967262,0.0002470612,0.6995323,0.007195883,0.1457987,0.1458724,0.0001201411],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01224227,0.001630826,0.8659545,0.001714912,0.0002872942,0.0002726821,0.005726524,0.1044629,0.007708177],"genre_scores_gemma":[0.1174141,0.00089794,0.8488277,0.001189232,0.00008124131,0.0003023394,0.01766166,0.008379709,0.005245999],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01748516,"threshold_uncertainty_score":0.03476679,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.008653647840289753,"score_gpt":0.2656501601318978,"score_spread":0.256996512291608,"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."}}