{"id":"W4413157588","doi":"10.1109/cvpr52734.2025.01239","title":"StoryGPT-V: Large Language Models as Consistent Story Visualizers","year":2025,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"Kootenay Association for Science & Technology","funders":"","keywords":"Computer science","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":[0.0003571525,0.0001097566,0.0001374389,0.0002186328,0.0001153693,0.00007270571,0.0006018012,0.00005671104,0.00006875706],"category_scores_gemma":[0.00002542068,0.0001022388,0.00006976179,0.000277489,0.00002140722,0.0002976446,0.0003273173,0.0001356929,0.00009307799],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006475265,"about_ca_system_score_gemma":0.0001373334,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001196453,"about_ca_topic_score_gemma":0.00003038079,"domain_scores_codex":[0.99892,0.00007976443,0.000153069,0.0003671104,0.0002101282,0.0002699912],"domain_scores_gemma":[0.9991171,0.00006831768,0.00003246434,0.0006633889,0.0000481973,0.00007053443],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000001984404,0.00004313499,0.00004563719,0.0000140581,0.00002136943,0.00001804767,0.00131984,0.0009680463,0.0003151791,0.9810347,0.01026318,0.005954837],"study_design_scores_gemma":[0.0006560556,0.00002602676,0.00006352133,0.00003420211,0.00001024552,0.000005318149,0.001075848,0.9502043,0.001145402,0.01341658,0.03311577,0.0002467957],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01506856,0.0005342668,0.8704867,0.001257126,0.0004786648,0.0001140679,9.602476e-7,0.0003246308,0.111735],"genre_scores_gemma":[0.9419132,0.00001589438,0.0278688,0.005193453,0.00002775542,0.00001457619,0.000001246122,0.000006584315,0.02495843],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9676181,"threshold_uncertainty_score":0.4169175,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01910971149770561,"score_gpt":0.2965980292981831,"score_spread":0.2774883178004775,"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."}}