{"id":"W4400267753","doi":"10.1145/3649217.3653554","title":"Can Small Language Models With Retrieval-Augmented Generation Replace Large Language Models When Learning Computer Science?","year":2024,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":21,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"Universitas Brawijaya","keywords":"Computer science; Language model; Natural language processing; Artificial intelligence; Universal Networking Language; Information retrieval; Natural language; Comprehension approach","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":[0.001155827,0.0002390715,0.0002037607,0.0003185882,0.000311113,0.001036208,0.0008678468,0.0000768049,0.00003613106],"category_scores_gemma":[0.00001813084,0.0001885685,0.00005630809,0.0007654267,0.00006109196,0.001465284,0.0004847184,0.0003505891,0.00002164076],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001984486,"about_ca_system_score_gemma":0.0003365297,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0004148103,"about_ca_topic_score_gemma":0.0002982565,"domain_scores_codex":[0.9972553,0.00008942989,0.0002959358,0.001074468,0.0006879185,0.000596959],"domain_scores_gemma":[0.9987447,0.00004813141,0.0000628604,0.0008311353,0.0001386389,0.0001745395],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00001675658,0.00005958359,0.00003529595,0.0000692432,0.00005099942,0.0002724793,0.04594856,0.6802015,0.0173389,0.2306816,0.0005303756,0.02479471],"study_design_scores_gemma":[0.0003006137,0.00008396312,0.000002743069,0.0000688041,0.00001012929,0.00004484949,0.0003241792,0.9915496,0.005757927,0.001473932,0.0001091958,0.0002740924],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1419319,0.0003622305,0.8528983,0.0009618772,0.0003021222,0.0002306535,0.000003353984,0.0008444003,0.002465139],"genre_scores_gemma":[0.7625313,0.000006597749,0.2330213,0.0004495664,0.0002243106,0.000007483416,0.00001384331,0.00002383441,0.003721746],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6205993,"threshold_uncertainty_score":0.9992179,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02659316822516559,"score_gpt":0.2482295856993994,"score_spread":0.2216364174742338,"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."}}