{"id":"W4389575929","doi":"10.1109/csnet59123.2023.10339778","title":"Stealthy Attacks on Multi-Agent Reinforcement Learning in Mobile Cyber-Physical Systems","year":2023,"lang":"en","type":"article","venue":"","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université Laval","funders":"National Science Foundation","keywords":"Computer science; Reinforcement learning; Adversarial system; Robustness (evolution); Scalability; Computer security; Distributed computing; Node (physics); Cyber-physical system; Artificial intelligence; Machine learning; Engineering","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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0007508108,0.0002096449,0.0002811845,0.0003073115,0.0001892762,0.0001440075,0.0007224083,0.00007228338,0.00001606694],"category_scores_gemma":[0.0001331519,0.0001869631,0.00007481342,0.0009552838,0.00002616099,0.000295044,0.0005811098,0.0005989805,0.001118996],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002284591,"about_ca_system_score_gemma":0.00006460882,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002462312,"about_ca_topic_score_gemma":0.00001212937,"domain_scores_codex":[0.9977211,0.0002620774,0.0003553567,0.0005584473,0.0005470405,0.0005559517],"domain_scores_gemma":[0.9988587,0.0003348738,0.000126799,0.0005165784,0.00004027141,0.0001227779],"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":[0.000006687444,0.00004874223,0.001341073,0.00002318488,0.000008026194,0.00004541611,0.001656776,0.9767554,0.00006352346,0.01285169,0.0002549754,0.006944545],"study_design_scores_gemma":[0.0006190864,0.0002410384,0.002024619,0.00005420802,0.000001924368,0.000002009408,0.0004633204,0.9908332,0.00005468646,0.00002285474,0.005470239,0.0002127789],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1756416,0.00002917976,0.8133087,0.0003951356,0.001314941,0.0008339669,3.430116e-7,0.001485707,0.006990444],"genre_scores_gemma":[0.9895465,0.00001262298,0.002088713,0.0001264867,0.0001380999,0.0001381163,0.000006904411,0.00002267199,0.00791988],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8139049,"threshold_uncertainty_score":0.9996588,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03201959749298665,"score_gpt":0.3216053526677072,"score_spread":0.2895857551747205,"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."}}