{"id":"W4405622294","doi":"10.48550/arxiv.2408.12109","title":"RoVRM: A Robust Visual Reward Model Optimized via Auxiliary Textual Preference Data","year":2024,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Image Retrieval and Classification Techniques","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Instituto de Ciencias del Mar y Limnología, Universidad Nacional Autónoma de México; Project 211; Natural Science Foundation of Liaoning Province; National Natural Science Foundation of China; Fundamental Research Funds for the Central Universities; Strong; Institute for Catastrophic Loss Reduction; National Science Foundation","keywords":"Preference; Computer science; Psychology; Artificial intelligence; Cognitive psychology; Economics; Microeconomics","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","open_science"],"consensus_categories":["open_science"],"category_scores_codex":[0.0005273064,0.0004899618,0.0004796605,0.0003710057,0.0001667751,0.0003950226,0.006241488,0.0004749443,0.00003259986],"category_scores_gemma":[0.0000610359,0.0005178178,0.0002095314,0.0008838327,0.0001921265,0.0007880931,0.01499843,0.001226139,0.0002418978],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002599281,"about_ca_system_score_gemma":0.0007920175,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001244472,"about_ca_topic_score_gemma":0.00000984587,"domain_scores_codex":[0.9963365,0.0001691043,0.0003700938,0.002430078,0.0002328952,0.0004613386],"domain_scores_gemma":[0.9956759,0.00009441151,0.0002687004,0.003467423,0.0002607628,0.0002328059],"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":[0.0004222821,0.001166998,0.0001031669,0.001247168,0.0007740186,0.001367583,0.0008221306,0.6329357,0.001398197,0.3003977,0.02981595,0.02954913],"study_design_scores_gemma":[0.000241455,0.00004190791,0.00001511383,0.000143033,0.0001080192,0.000007768303,0.00002549314,0.9402882,0.0006439827,0.05762778,0.0003019234,0.0005553287],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001766389,0.0001446628,0.9918041,0.0005007493,0.0003163168,0.0004768314,0.0001260763,0.001470746,0.003394108],"genre_scores_gemma":[0.9261934,0.0004536515,0.06307129,0.0001259888,0.000118142,0.000004572839,0.0002183689,0.00004144884,0.009773158],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9287328,"threshold_uncertainty_score":0.9997274,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2234317025266224,"score_gpt":0.2407001204958311,"score_spread":0.01726841796920861,"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."}}