{"id":"W3197018685","doi":"","title":"Where Did You Learn That From? Surprising Effectiveness of Membership Inference Attacks Against Temporally Correlated Data in Deep Reinforcement Learning.","year":2021,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo; McGill University","funders":"","keywords":"Reinforcement learning; Adversarial system; Artificial intelligence; Reinforcement; Computer science; Deep learning; Inference; Machine learning; Vulnerability (computing); Computer security; Psychology; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.006515538,0.0006005618,0.0006546215,0.0003856775,0.0008054061,0.001093891,0.001168455,0.001746237,0.001464567],"category_scores_gemma":[0.04564962,0.00038969,0.0007155232,0.0003348716,0.002737417,0.003460367,0.002502341,0.004240634,0.000311714],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001475818,"about_ca_system_score_gemma":0.001321654,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002344916,"about_ca_topic_score_gemma":0.002147947,"domain_scores_codex":[0.9953041,0.002462465,0.0001485602,0.0007484343,0.0009688458,0.0003675532],"domain_scores_gemma":[0.9666249,0.0252468,0.002218414,0.004157695,0.00102507,0.0007270786],"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":[0.002284363,0.0005883544,0.01879564,0.0002597086,0.0003702795,0.0006696936,0.0007834098,0.735513,0.01236146,0.09593848,0.008483546,0.1239522],"study_design_scores_gemma":[0.00004746731,0.0001189888,0.001097656,0.00002390868,0.00001916619,0.00007841681,0.00004466862,0.9476273,0.00419074,0.04606578,0.0006676071,0.00001831157],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4274077,0.0008098293,0.557197,0.005270959,0.0002221412,0.0001510006,0.0003896662,0.001252591,0.0072992],"genre_scores_gemma":[0.9768723,0.00009381372,0.0213308,0.0003687032,0.00002402647,0.00004674727,0.000114414,0.00005062198,0.001098459],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006515538,"threshold_uncertainty_score":0.03445786,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08241447342643901,"score_gpt":0.2403000220203051,"score_spread":0.1578855485938661,"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."}}