{"id":"W4415306945","doi":"10.1109/iccv51701.2025.00863","title":"Perspective-Aware Reasoning in Vision-Language Models via Mental Imagery Simulation","year":2025,"lang":"en","type":"preprint","venue":"","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Kootenay Association for Science & Technology","funders":"","keywords":"Perspective (graphical); Visual reasoning; Focus (optics); Construct (python library); Bridge (graph theory); Mental image; Spatial intelligence; Orientation (vector space); Object (grammar)","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"],"consensus_categories":[],"category_scores_codex":[0.0003719918,0.0003080773,0.0003152209,0.000468305,0.0001367242,0.0002722793,0.001225564,0.0002213912,0.00004150414],"category_scores_gemma":[0.00008700489,0.0003216567,0.0001303366,0.0004743319,0.00003189109,0.0004327962,0.002479036,0.0009437059,0.00003701013],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006350387,"about_ca_system_score_gemma":0.0001875582,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009461692,"about_ca_topic_score_gemma":0.0001265603,"domain_scores_codex":[0.997672,0.0001740104,0.0003866357,0.001119192,0.0003580906,0.0002900537],"domain_scores_gemma":[0.9980833,0.0002628948,0.0001808971,0.001236927,0.000158845,0.00007714242],"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":[0.000004761203,0.00009864567,0.0003383027,0.00003159217,0.00001479365,0.000006052295,0.005105718,0.9555285,0.0000888288,0.01879365,0.00005816996,0.01993096],"study_design_scores_gemma":[0.0002074597,0.000009288062,0.004953718,0.0001676218,0.000005599506,0.000001583613,0.000205352,0.9773979,0.00004689099,0.01670032,0.00001395198,0.0002903173],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01370415,0.0001687238,0.9675068,0.001529211,0.000223493,0.0006848962,0.00001691324,0.0005358855,0.01562987],"genre_scores_gemma":[0.8783432,0.000007222307,0.1206929,0.0001680956,0.00004750854,0.0001066667,0.00004062683,0.00001437007,0.0005794202],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.864639,"threshold_uncertainty_score":0.9999235,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01329232318466052,"score_gpt":0.3496713324522095,"score_spread":0.336379009267549,"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."}}