{"id":"W4412388539","doi":"10.1007/978-3-031-85240-4_14","title":"Detecting Jailbreaking Prompts: an Anti-Persuasion Filter Framework","year":2025,"lang":"en","type":"book-chapter","venue":"Lecture notes in social networks","topic":"Deception detection and forensic psychology","field":"Psychology","cited_by":2,"is_retracted":false,"has_abstract":false,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"","keywords":"Persuasion; Computer science; Psychology; Human–computer interaction; Communication; Social psychology","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","insufficient_payload"],"consensus_categories":["research_integrity"],"category_scores_codex":[0.0004423945,0.0006633949,0.0008586104,0.0004574108,0.0004531875,0.00008564446,0.00042127,0.004001495,0.007478322],"category_scores_gemma":[0.0001390628,0.0007013132,0.0003620475,0.0003006184,0.0002820677,0.00006898499,0.00012098,0.003705201,0.0001376378],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001786067,"about_ca_system_score_gemma":0.00003670679,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000296679,"about_ca_topic_score_gemma":0.0001522268,"domain_scores_codex":[0.9968433,0.0002439953,0.0007288025,0.00114952,0.0002692967,0.0007650563],"domain_scores_gemma":[0.9981274,0.0005534858,0.0004513948,0.0006522957,0.0001078914,0.0001075204],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0002823543,0.00008670658,0.0002874784,0.00002384251,0.0002574127,0.0001088841,0.003586337,0.0004269153,0.00001218284,0.02453758,0.003116909,0.9672734],"study_design_scores_gemma":[0.004662918,0.001094234,0.008267669,0.00254586,0.0006244265,0.0003098797,0.0003533083,0.0032703,0.00004661846,0.5546477,0.4191478,0.005029282],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"other","genre_gemma":"empirical","genre_scores_codex":[0.001074764,0.00160864,0.4151189,0.000642686,0.01531416,0.0009236877,0.00002672305,0.0005797841,0.5647106],"genre_scores_gemma":[0.9503542,0.0001059714,0.00300569,0.007139961,0.01156835,0.0001003689,0.0001651245,0.0002814916,0.02727883],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9622441,"threshold_uncertainty_score":0.9995438,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02999823860595933,"score_gpt":0.33592303416166,"score_spread":0.3059247955557006,"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."}}