{"id":"W4410235858","doi":"10.3390/jcp5020023","title":"Combining Supervised and Reinforcement Learning to Build a Generic Defensive Cyber Agent","year":2025,"lang":"en","type":"article","venue":"Journal of Cybersecurity and Privacy","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"Carleton University","funders":"","keywords":"Reinforcement learning; Reinforcement; Computer science; Artificial intelligence; 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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006785452,0.0001749225,0.0003242323,0.0003165535,0.0002504875,0.0002960606,0.0004968419,0.00007187952,0.00001112124],"category_scores_gemma":[0.0004073135,0.0001600621,0.00008141954,0.0003766444,0.00004985511,0.000478992,0.000716497,0.0004769652,0.000004538335],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006633619,"about_ca_system_score_gemma":0.0001138769,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002041854,"about_ca_topic_score_gemma":0.000001297973,"domain_scores_codex":[0.9984765,0.0001215693,0.0005416075,0.0002171144,0.000360889,0.0002823237],"domain_scores_gemma":[0.9987898,0.0002033975,0.0002917475,0.0002578055,0.0002419176,0.0002152738],"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":[0.0002478646,0.0001202408,0.02465175,0.000309511,0.0006491392,0.0002816767,0.07230311,0.7513573,0.003171063,0.08942847,0.006203859,0.05127603],"study_design_scores_gemma":[0.008505552,0.003520338,0.0357253,0.001807472,0.0003373107,0.0006460349,0.003637445,0.6262958,0.003577279,0.007210657,0.3072114,0.001525399],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3912361,0.0006261439,0.6029814,0.002728935,0.0004351866,0.000177003,1.298182e-7,0.00003031999,0.001784809],"genre_scores_gemma":[0.9737915,0.0004035947,0.02373914,0.00157697,0.00005123525,0.000002140661,5.951337e-7,0.000007264971,0.0004275584],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.5825554,"threshold_uncertainty_score":0.6527145,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01741083748785987,"score_gpt":0.2618763526690385,"score_spread":0.2444655151811786,"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."}}