{"id":"W4410537502","doi":"10.1145/3736407","title":"CodeUltraFeedback: An LLM-as-a-Judge Dataset for Aligning Large Language Models to Coding Preferences","year":2025,"lang":"en","type":"article","venue":"ACM Transactions on Software Engineering and Methodology","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"","keywords":"Computer science; Coding (social sciences); Natural language processing; Artificial intelligence; Programming language; Statistics; Mathematics","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":[0.0008498564,0.0001909016,0.0002645285,0.0003342133,0.0001896142,0.00009191851,0.0007408879,0.0001480191,0.000003319322],"category_scores_gemma":[0.0006868558,0.0001842456,0.00004157644,0.0003415741,0.00001743926,0.0004074826,0.00004382261,0.0002455781,0.000001276155],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003093448,"about_ca_system_score_gemma":0.00004650041,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00005863178,"about_ca_topic_score_gemma":0.00002267769,"domain_scores_codex":[0.9987627,0.0001118598,0.0001948747,0.0004987453,0.0000925511,0.0003392631],"domain_scores_gemma":[0.9976281,0.001568841,0.00003391874,0.0006042821,0.00005772377,0.0001071223],"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":[0.0002154193,0.0003685963,0.00002660224,0.0008130405,0.0002939534,0.00002906472,0.01490711,0.02378146,0.05436829,0.153832,0.0009888462,0.7503756],"study_design_scores_gemma":[0.00272955,0.002179902,0.00008468506,0.001011563,0.0002985027,0.0001805385,0.001752785,0.2052687,0.6252098,0.1444063,0.0145281,0.002349588],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.002432794,0.0005158159,0.9951476,0.0002606182,0.0003295244,0.0002522442,0.0002701776,0.0007869856,0.000004216465],"genre_scores_gemma":[0.08475281,0.00002781812,0.9142609,0.0006654977,0.00002154598,0.0001223264,0.00007131341,0.00001435343,0.00006344233],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.748026,"threshold_uncertainty_score":0.7513319,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0823360169353762,"score_gpt":0.3675850040090939,"score_spread":0.2852489870737177,"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."}}