{"id":"W4409381863","doi":"10.1609/aaai.v39i28.35158","title":"Exploring the Efficacy of Multi-Agent Reinforcement Learning for Autonomous Cyber Defence: A CAGE Challenge 4 Perspective","year":2025,"lang":"en","type":"article","venue":"Proceedings of the AAAI Conference on Artificial Intelligence","topic":"Advanced Malware Detection Techniques","field":"Computer Science","cited_by":10,"is_retracted":false,"has_abstract":true,"ca_institutions":"Defence Research and Development Canada","funders":"","keywords":"Perspective (graphical); Reinforcement learning; Reinforcement; Cage; Computer science; Psychology; Cognitive science; Human–computer interaction; Engineering; Social psychology; Artificial intelligence; Structural 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":[],"consensus_categories":[],"category_scores_codex":[0.0006033244,0.0002303324,0.0002801353,0.0001956215,0.0003690379,0.0001055709,0.001785824,0.00005907896,0.000007592304],"category_scores_gemma":[0.001304795,0.0001619606,0.0001929457,0.0006816479,0.0002759478,0.0004432177,0.0005403726,0.0003740668,0.00000595302],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001883738,"about_ca_system_score_gemma":0.0001284843,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000870962,"about_ca_topic_score_gemma":0.00001476461,"domain_scores_codex":[0.9982064,0.00002367068,0.0005933753,0.0005097729,0.0003390392,0.0003277654],"domain_scores_gemma":[0.9976067,0.0003139953,0.0004873731,0.0003669553,0.00118102,0.00004397989],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0000846343,0.0001293161,0.00001448509,0.00005734704,0.00003180161,1.501007e-7,0.00640169,0.002885558,0.01475802,0.8951551,0.00001690312,0.08046503],"study_design_scores_gemma":[0.00007363049,0.0003739721,0.0001275409,0.0003333536,0.00002091458,0.000001089952,0.003771213,0.1008227,0.8390654,0.05486751,0.0003515245,0.0001911181],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02839059,0.00006297065,0.9615685,0.003081638,0.0005199508,0.001534355,0.000001495103,0.0002512172,0.004589225],"genre_scores_gemma":[0.9904304,0.0001170256,0.008533916,0.00007745479,0.000029708,0.0003607599,1.516899e-7,0.00001167257,0.0004389362],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9620398,"threshold_uncertainty_score":0.660456,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1502720827387915,"score_gpt":0.3429535311925003,"score_spread":0.1926814484537089,"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."}}