{"id":"W7117557791","doi":"10.5281/zenodo.18080355","title":"The Instruction Stack Audit Framework (ISAF): A Technical Methodology for Tracing AI Accountability Across Nine Abstraction Layers","year":2025,"lang":"","type":"preprint","venue":"Zenodo (CERN European Organization for Nuclear Research)","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Accountability; Audit; Corporate governance; Documentation; Transparency (behavior); Abstraction; Traceability; Protocol stack","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch","metaepi_narrow","sts","scholarly_communication","open_science","research_integrity","insufficient_payload"],"consensus_categories":["open_science"],"category_scores_codex":[0.01284596,0.0007508154,0.000836648,0.0003945647,0.01518651,0.005217928,0.006581015,0.001118934,0.001751932],"category_scores_gemma":[0.02122504,0.0007466545,0.0004568313,0.001795072,0.001185712,0.00119649,0.01234818,0.005519661,0.0005928352],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001749094,"about_ca_system_score_gemma":0.0001061673,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001514747,"about_ca_topic_score_gemma":0.00001547406,"domain_scores_codex":[0.9893121,0.003868916,0.001605035,0.002526517,0.001161481,0.001525934],"domain_scores_gemma":[0.9902017,0.002418003,0.001327937,0.002812342,0.002909298,0.0003307616],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.001337496,0.0003270553,0.00002707017,0.0009211955,0.0003727431,0.000009578202,0.006676677,0.09582593,0.0009453922,0.04211623,0.01329463,0.838146],"study_design_scores_gemma":[0.001829567,0.000680586,0.003921477,0.000605541,0.0001793301,0.0003202847,0.002034039,0.09647794,0.000511157,0.01967868,0.8725597,0.00120174],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.005035621,0.0001248583,0.9755595,0.007021776,0.003684734,0.002658592,0.0004236803,0.001544269,0.003946934],"genre_scores_gemma":[0.7958503,0.0005839838,0.193698,0.000522522,0.002708695,0.000005334398,0.00178626,0.002879742,0.001965101],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.859265,"threshold_uncertainty_score":0.9994984,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08141551385463358,"score_gpt":0.3767917135260739,"score_spread":0.2953761996714403,"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."}}