{"id":"W4416018545","doi":"10.48550/arxiv.2511.03112","title":"Accelerating inverse materials design using generative diffusion models with reinforcement learning","year":2025,"lang":"","type":"preprint","venue":"ArXiv.org","topic":"Machine Learning in Materials Science","field":"Materials Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung; National Science Foundation","keywords":"Property (philosophy); Computation; Workflow; Reinforcement learning; Inverse; Diffusion","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","sts","scholarly_communication","insufficient_payload"],"consensus_categories":["metaepi_narrow"],"category_scores_codex":[0.006525216,0.002076637,0.002357984,0.0006553028,0.003259381,0.002484553,0.002667415,0.0009246311,0.00470421],"category_scores_gemma":[0.001055936,0.00183641,0.0002387129,0.000924218,0.0009210704,0.001896886,0.005642514,0.001761668,0.0003661176],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001113165,"about_ca_system_score_gemma":0.002186648,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002848569,"about_ca_topic_score_gemma":0.00004780158,"domain_scores_codex":[0.9848056,0.004404775,0.002897835,0.003733123,0.00192439,0.002234262],"domain_scores_gemma":[0.9924529,0.0005852861,0.003444791,0.002076049,0.0009310955,0.0005098963],"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":[0.000340446,0.00006220325,0.003059745,0.0004044552,0.00005404425,0.00003677419,0.00254732,0.5139957,0.4792217,0.0002066821,0.00002010997,0.00005085584],"study_design_scores_gemma":[0.0009380677,0.0004103942,0.0005350252,0.002293911,0.0002371891,0.00002629873,0.0004619421,0.5882838,0.4050971,0.0002515363,0.00005770507,0.001407051],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6267738,0.00005134933,0.3671316,0.0001226935,0.002813712,0.002047157,0.000030702,0.0002836924,0.0007452461],"genre_scores_gemma":[0.8192662,0.0002436176,0.1750344,0.000499803,0.0007554466,0.0002940541,0.00009718326,0.0001473428,0.003661903],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1924923,"threshold_uncertainty_score":0.9991975,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.122121986936091,"score_gpt":0.3095336895479037,"score_spread":0.1874117026118127,"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."}}