{"id":"W7102329764","doi":"","title":"Offline Preference Optimization via Maximum Marginal Likelihood Estimation","year":2025,"lang":"","type":"article","venue":"ArXiv.org","topic":"Recommender Systems and Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Hyperparameter; Preference; Marginal likelihood; Entropy (arrow of time); Principle of maximum entropy; Revealed preference","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":[0.002739063,0.001060648,0.001550369,0.0006701582,0.0005353841,0.001512809,0.001747067,0.001309059,0.005117416],"category_scores_gemma":[0.01429691,0.0008397043,0.0009317021,0.001029354,0.0007830846,0.002613281,0.001610298,0.002519453,0.002685882],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009613073,"about_ca_system_score_gemma":0.001635254,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00372485,"about_ca_topic_score_gemma":0.00654533,"domain_scores_codex":[0.9978033,0.00112176,0.0001029857,0.0003912989,0.0004118559,0.0001687797],"domain_scores_gemma":[0.9950866,0.003239036,0.0002827253,0.0006905675,0.0005271628,0.0001739388],"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":[0.0006012816,0.0003419906,0.004797554,0.0003522031,0.0001988861,0.0002469831,0.0003879958,0.4927891,0.01076363,0.02772042,0.01001123,0.4517886],"study_design_scores_gemma":[0.00002522345,0.00005073074,0.0002729247,0.000008089484,0.0000108202,0.00004510447,0.00003004055,0.9841262,0.001569015,0.01307554,0.0007728204,0.00001349468],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01816907,0.0001840159,0.9776593,0.0003592711,0.00003503532,0.00006913758,0.0001108918,0.00162399,0.001789391],"genre_scores_gemma":[0.6079624,0.0001656248,0.3845328,0.0004800806,0.0001128922,0.000237543,0.0005686702,0.0005145044,0.005425529],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005117416,"threshold_uncertainty_score":0.01711947,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03695703122362177,"score_gpt":0.2691590736295219,"score_spread":0.2322020424059001,"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."}}