{"id":"W4409362363","doi":"10.1609/aaai.v39i25.34821","title":"Perception-Guided Jailbreak Against Text-to-Image Models","year":2025,"lang":"en","type":"article","venue":"Proceedings of the AAAI Conference on Artificial Intelligence","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":10,"is_retracted":false,"has_abstract":true,"ca_institutions":"Ministry of Education and Child Care","funders":"Fundamental Research Funds for the Central Universities; Ministry of Education of the People's Republic of China; National Natural Science Foundation of China; National Research Foundation Singapore; National Research Foundation","keywords":"Perception; Image (mathematics); Computer science; Psychology; Artificial intelligence; Cognitive psychology; Natural language processing; Neuroscience","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":[0.0007147523,0.0003118135,0.0003487593,0.00033538,0.0003837862,0.0004371933,0.0036231,0.0001258827,0.00005595271],"category_scores_gemma":[0.0009304992,0.0002527052,0.0001643451,0.001507774,0.0002549913,0.00082288,0.001194708,0.0005192498,0.0001927648],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001265642,"about_ca_system_score_gemma":0.0001856524,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006585648,"about_ca_topic_score_gemma":0.000006103941,"domain_scores_codex":[0.9974719,0.00003738501,0.000682355,0.0007579043,0.0005839583,0.0004664657],"domain_scores_gemma":[0.9979605,0.0001389936,0.0003011946,0.0005425321,0.0009371658,0.0001196091],"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":[0.00003268176,0.00008278508,0.00008769549,0.0000356868,0.00001590317,5.051576e-7,0.001510864,0.009146844,0.04194717,0.8578955,0.0007398252,0.08850454],"study_design_scores_gemma":[0.00004135526,0.00007285838,0.0002410005,0.0003780186,0.00001450126,0.000002012218,0.0009318044,0.6590501,0.06666764,0.2721128,0.0001782586,0.0003096233],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0600111,0.00000779116,0.8099911,0.0128012,0.0007806814,0.0006438217,0.000003179914,0.0002497364,0.1155115],"genre_scores_gemma":[0.9611923,0.00001663137,0.03633685,0.001203933,0.00006930654,0.00004010189,4.891398e-7,0.00001573338,0.001124641],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9011812,"threshold_uncertainty_score":0.9999925,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05902485939270401,"score_gpt":0.3197705640422241,"score_spread":0.2607457046495201,"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."}}