{"id":"W4413217936","doi":"10.2139/ssrn.5386966","title":"A Reward-Directed Diffusion Framework for Generative Design Optimization","year":2025,"lang":"en","type":"preprint","venue":"SSRN Electronic Journal","topic":"Advanced Multi-Objective Optimization Algorithms","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of British Columbia; University of British Columbia Hospital","funders":"","keywords":"Diffusion; Generative grammar; Computer science; Mathematical optimization; Mathematics; Artificial intelligence; Thermodynamics; Physics","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":["metaepi_narrow","research_integrity"],"consensus_categories":[],"category_scores_codex":[0.00162837,0.0005526778,0.0005801003,0.0005327397,0.000646193,0.0004313362,0.001636467,0.0005835986,0.00001299713],"category_scores_gemma":[0.001049525,0.0005424799,0.0003110874,0.0007119672,0.00005287902,0.0004573085,0.0007810649,0.00415169,0.000003596773],"about_ca_system_candidate":true,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003039313,"about_ca_system_score_gemma":0.0072894,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001015892,"about_ca_topic_score_gemma":0.00001799721,"domain_scores_codex":[0.9951274,0.000512914,0.0006779914,0.001030133,0.0004734349,0.002178063],"domain_scores_gemma":[0.9965359,0.0005346446,0.0008552871,0.0007579063,0.001173088,0.0001431989],"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":[0.00006068435,0.00009548832,0.000003889856,0.00001616416,0.0002042534,0.000002157889,0.0003581461,0.8370576,0.00001395507,0.1356443,0.00006882628,0.0264746],"study_design_scores_gemma":[0.0005293591,0.0001443264,0.000002685397,0.0001356677,0.00003796951,0.00003628382,0.00005955585,0.5869858,0.0001539003,0.4115055,0.00008772781,0.0003212142],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.000008614706,0.00208989,0.9929494,0.001084058,0.001797328,0.001465433,0.00001967557,0.0004725919,0.0001129841],"genre_scores_gemma":[0.0005165354,0.006767021,0.9898013,0.0002588271,0.000497834,0.0002536338,0.00004544791,0.00005090193,0.001808456],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.2758612,"threshold_uncertainty_score":0.9997027,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01965408554451329,"score_gpt":0.2951295512575384,"score_spread":0.2754754657130251,"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."}}