{"id":"W4415589112","doi":"10.1145/3773034","title":"Synthesizing Efficient and Permissive Programmatic Runtime Shields for Neural Policies","year":2025,"lang":"en","type":"article","venue":"ACM Transactions on Software Engineering and Methodology","topic":"Neural Networks and Applications","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Alberta Medical Association","funders":"","keywords":"Shields; Overhead (engineering); Reduction (mathematics); Software; Control (management); Reliability (semiconductor); Safety standards; Runtime system","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001137354,0.001124069,0.0005875365,0.000617406,0.0004250744,0.0008052084,0.0009090802,0.000806218,0.004091311],"category_scores_gemma":[0.005476373,0.0005283346,0.001064096,0.0002406192,0.001362902,0.001182854,0.001410154,0.001399534,0.0006805781],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009396314,"about_ca_system_score_gemma":0.002551597,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002903355,"about_ca_topic_score_gemma":0.004991282,"domain_scores_codex":[0.9990644,0.0002198848,0.00005884279,0.000169978,0.0003302716,0.0001566272],"domain_scores_gemma":[0.9978811,0.001303313,0.0001897642,0.0003250522,0.0002247246,0.00007605667],"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":[0.0002568699,0.0001230542,0.003080454,0.0005542826,0.00006441872,0.000284458,0.0003428392,0.7486378,0.03915267,0.0331742,0.003239276,0.1710897],"study_design_scores_gemma":[0.00003236884,0.00009522573,0.0001972077,0.00003418319,0.00002486053,0.00005610688,0.00004428108,0.9719136,0.01401785,0.01018969,0.003380838,0.00001371524],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.06209699,0.0004412193,0.9226781,0.0002990896,0.00008241509,0.0001863702,0.0002219329,0.007832948,0.00616087],"genre_scores_gemma":[0.5953451,0.0003534537,0.3994128,0.0002562289,0.00002989796,0.0003212597,0.0005069383,0.001239341,0.002535086],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004091311,"threshold_uncertainty_score":0.01368684,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0516818758031873,"score_gpt":0.3185185259233981,"score_spread":0.2668366501202108,"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."}}