{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0003184528,0.0001238637,0.000181687,0.0001552162,0.00007159523,0.00007574625,0.0008714878,0.00005689188,0.000001167624],"category_scores_gemma":[0.0001202365,0.00008305855,0.00005393431,0.0004496788,0.00008274725,0.0004915234,0.0006316694,0.0001345764,1.92348e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005523765,"about_ca_system_score_gemma":0.00002876686,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001176784,"about_ca_topic_score_gemma":0.000008585963,"domain_scores_codex":[0.9990987,0.000007502152,0.0002669253,0.000245603,0.0002476933,0.0001335944],"domain_scores_gemma":[0.9993179,0.00003537786,0.0002250534,0.0001776999,0.0002185749,0.00002541967],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.00001158498,0.00004561285,0.000171787,0.0002194314,0.00002051892,2.420845e-7,0.007992242,0.000005333914,0.3429043,0.5893523,0.0002517776,0.05902484],"study_design_scores_gemma":[0.0001331252,0.00002952002,0.00004569668,0.0001844373,0.00001227804,0.000002515751,0.0003895514,0.009148172,0.6675795,0.3223919,0.00001839454,0.00006492574],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4490298,0.01663191,0.5080837,0.006982036,0.00049778,0.001475221,0.000003618203,0.0005973258,0.01669862],"genre_scores_gemma":[0.8198452,0.00003074143,0.1798193,0.0001446939,0.00001035277,0.00001935091,3.256705e-7,0.000004554774,0.0001254893],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.3708154,"threshold_uncertainty_score":0.3387029,"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."}}