{"id":"W7131116583","doi":"10.1109/icairc68035.2025.11384878","title":"Two-Stage Fine-Tuning for Retrieval-Augmented Large Language Models via Retriever and Prompt Optimization","year":2025,"lang":"","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Language model; Generalization; Generator (circuit theory); Labrador Retriever; Protocol (science); Set (abstract data type)","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002820651,0.002133547,0.001582299,0.001030794,0.0005144554,0.001654453,0.002795942,0.001868452,0.01124383],"category_scores_gemma":[0.01078605,0.000833737,0.001695845,0.0007137539,0.0009733638,0.002770269,0.003035455,0.003383623,0.008090205],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007160843,"about_ca_system_score_gemma":0.001365422,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002173545,"about_ca_topic_score_gemma":0.004927201,"domain_scores_codex":[0.9981965,0.0007424817,0.0001072212,0.0005466051,0.0002564416,0.0001508015],"domain_scores_gemma":[0.9970732,0.001765217,0.0001043749,0.0006438376,0.0002994834,0.0001138302],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0009745896,0.0006073444,0.002832095,0.0008824416,0.0002947176,0.0006273407,0.0006304078,0.2010508,0.04919706,0.01129931,0.03796227,0.6936415],"study_design_scores_gemma":[0.0002517213,0.0002408973,0.000501582,0.00004220303,0.0000943487,0.0002985777,0.0001049759,0.9591458,0.01729834,0.01354478,0.00841693,0.00005981175],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.02504213,0.001073915,0.9176107,0.0004740781,0.0002505706,0.0003957095,0.0005636218,0.05116389,0.003425335],"genre_scores_gemma":[0.4068992,0.0005067031,0.5716441,0.001514498,0.000260426,0.001103617,0.003740808,0.006119174,0.008211453],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01124383,"threshold_uncertainty_score":0.03761435,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02058260023157539,"score_gpt":0.2826260557435897,"score_spread":0.2620434555120143,"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."}}