{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,15]],"date-time":"2026-07-15T16:10:28Z","timestamp":1784131828585,"version":"3.55.0"},"reference-count":38,"publisher":"Elsevier BV","license":[{"start":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T00:00:00Z","timestamp":1788220800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/tdm\/userlicense\/1.0\/"},{"start":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T00:00:00Z","timestamp":1788220800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/legal\/tdmrep-license"},{"start":{"date-parts":[[2026,6,24]],"date-time":"2026-06-24T00:00:00Z","timestamp":1782259200000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["elsevier.com","sciencedirect.com"],"crossmark-restriction":true},"short-container-title":["Journal of Information Security and Applications"],"published-print":{"date-parts":[[2026,9]]},"DOI":"10.1016\/j.jisa.2026.104507","type":"journal-article","created":{"date-parts":[[2026,6,2]],"date-time":"2026-06-02T22:02:51Z","timestamp":1780437771000},"page":"104507","update-policy":"https:\/\/doi.org\/10.1016\/elsevier_cm_policy","source":"Crossref","is-referenced-by-count":1,"special_numbering":"C","title":["Unveiling the black box: A multi-layer framework for explaining reinforcement learning-based cyber agents"],"prefix":"10.1016","volume":"101","author":[{"given":"Diksha","family":"Goel","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kristen","family":"Moore","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jeff","family":"Wang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6161-7982","authenticated-orcid":false,"given":"Minjune","family":"Kim","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9709-1663","authenticated-orcid":false,"given":"Thanh Thi","family":"Nguyen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","reference":[{"key":"10.1016\/j.jisa.2026.104507_bib0001","series-title":"Proceedings of the genetic and evolutionary computation conference","first-page":"1191","article-title":"Defending active directory by combining neural network based dynamic program and evolutionary diversity optimisation","author":"Goel","year":"2022"},{"key":"10.1016\/j.jisa.2026.104507_bib0002","unstructured":"Microsoft Defender Research TeamCyberbattlesim. https:\/\/github.com\/microsoft\/cyberbattlesim; 2021. Created by Christian Seifert, Michael Betser, William Blum, James Bono, Kate Farris, Emily Goren, Justin Grana, Kristian Holsheimer, Brandon Marken, Joshua Neil, Nicole Nichols, Jugal Parikh, Haoran Wei."},{"key":"10.1016\/j.jisa.2026.104507_bib0003","unstructured":"Cyber operations research gym. https:\/\/github.com\/cage-challenge\/CybORG; 2022. Created by Maxwell Standen, David Bowman, Son Hoang, Toby Richer, Martin Lucas, Richard Van Tassel, Phillip Vu, Mitchell Kiely, KC C., Natalie Konschnik, Joshua Collyer."},{"key":"10.1016\/j.jisa.2026.104507_bib0004","doi-asserted-by":"crossref","unstructured":"Terranova F., Lahmadi A., Chrisment I.. Scalable and generalizable RL agents for attack path discovery via continuous invariant spaces. In: 2025 28th International symposium on research in attacks, intrusions and defenses (RAID). Gold Coast, Australia; 2025, p. 18. https:\/\/hal.science\/hal-05182437.","DOI":"10.1109\/RAID67961.2025.00029"},{"key":"10.1016\/j.jisa.2026.104507_bib0005","series-title":"Proceedings of the genetic and evolutionary computation conference","first-page":"1348","article-title":"Evolving reinforcement learning environment to minimize learner\u2019s achievable reward: an application on hardening active directory systems","author":"Goel","year":"2023"},{"key":"10.1016\/j.jisa.2026.104507_bib0006","unstructured":"Goel D.. Enhancing network resilience through machine learning-powered graph combinatorial optimization: applications in cyber defense and information diffusion, 2023. arXiv: 2310.10667."},{"key":"10.1016\/j.jisa.2026.104507_bib0007","series-title":"Conference on learning theory","first-page":"4300","article-title":"Non-stationary reinforcement learning without prior knowledge: an optimal black-box approach","author":"Wei","year":"2021"},{"issue":"9","key":"10.1016\/j.jisa.2026.104507_bib0008","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3561048","article-title":"Explainable AI (XAI): core ideas, techniques, and solutions","volume":"55","author":"Dwivedi","year":"2023","journal-title":"ACM Comput Surv"},{"key":"10.1016\/j.jisa.2026.104507_bib0009","doi-asserted-by":"crossref","unstructured":"Gyevnar B., Wang C., Lucas C.G., Cohen S.B., Albrecht S.V.. Causal explanations for sequential decision-making in multi-agent systems, 2023. arXiv: 2302.10809.","DOI":"10.65109\/FHAA7375"},{"key":"10.1016\/j.jisa.2026.104507_bib0010","unstructured":"Mathes T.K., Inman J., Col\u00f3n A., Khan S.. CODEX: A cluster-based method for explainable reinforcement learning, 2023. arXiv: 2312.04216."},{"key":"10.1016\/j.jisa.2026.104507_bib0011","series-title":"Proceedings of the AAAI conference on artificial intelligence","first-page":"2493","article-title":"Explainable reinforcement learning through a causal lens","volume":"vol. 34","author":"Madumal","year":"2020"},{"key":"10.1016\/j.jisa.2026.104507_bib0012","doi-asserted-by":"crossref","DOI":"10.1613\/jair.1.18126","article-title":"Causal explanations for sequential decision making","volume":"83","author":"Nashed","year":"2025","journal-title":"J Artif Intell Res"},{"key":"10.1016\/j.jisa.2026.104507_bib0013","series-title":"32nd USENIX security symposium (USENIX security 23)","first-page":"7375","article-title":"AIRS: explanation for deep reinforcement learning-based security applications","author":"Yu","year":"2023"},{"key":"10.1016\/j.jisa.2026.104507_bib0014","unstructured":"Foley M., Wang M., Hicks C., Mavroudis V., et al. Inroads into autonomous network defence using explained reinforcement learning, 2023. arXiv: 2306.09318."},{"key":"10.1016\/j.jisa.2026.104507_bib0015","unstructured":"Alabdulkarim A., Singh M., Mansi G., Hall K., Riedl M.O.. Experiential explanations for reinforcement learning, 2022. arXiv: 2210.04723."},{"key":"10.1016\/j.jisa.2026.104507_bib0016","doi-asserted-by":"crossref","DOI":"10.1016\/j.compeleceng.2022.108356","article-title":"Explainable artificial intelligence for cybersecurity","volume":"103","author":"Sharma","year":"2022","journal-title":"Comput Electr Eng"},{"key":"10.1016\/j.jisa.2026.104507_bib0017","series-title":"Proceedings of the FPT AI conference","first-page":"1","article-title":"Evaluation of explainable artificial intelligence: shap, lime, and cam","author":"Nguyen","year":"2021"},{"key":"10.1016\/j.jisa.2026.104507_bib0018","series-title":"2021 IEEE Symposium series on computational intelligence (SSCI)","first-page":"01","article-title":"Explainability of cybersecurity threats data using shap","author":"Alenezi","year":"2021"},{"key":"10.1016\/j.jisa.2026.104507_bib0019","unstructured":"Claypoole J., Cheung S., Gehani A., Yegneswaran V., Ridley A.. Interpreting agent behaviors in reinforcement-learning-based cyber-battle simulation platforms, 2025. arXiv: 2506.08192."},{"key":"10.1016\/j.jisa.2026.104507_bib0020","unstructured":"Schwartz J., Kurniawatti H.. Nasim: network attack simulator, 2019. https:\/\/networkattacksimulator.readthedocs.io\/."},{"key":"10.1016\/j.jisa.2026.104507_bib0021","unstructured":"Molina-Markham A., Winder R.K., Ridley A.. Network defense is not a game, 2021. arXiv: 2104.10262."},{"key":"10.1016\/j.jisa.2026.104507_bib0022","series-title":"Proceedings of the workshop on autonomous cybersecurity","first-page":"56","article-title":"Entity-based reinforcement learning for autonomous cyber defence","author":"Thompson","year":"2024"},{"key":"10.1016\/j.jisa.2026.104507_bib0023","series-title":"European symposium on research in computer security","first-page":"332","article-title":"Optimizing cyber defense in dynamic active directories through reinforcement learning","author":"Goel","year":"2024"},{"key":"10.1016\/j.jisa.2026.104507_bib0024","unstructured":"Wiebe J., Mallah R.A., Li L.. Learning cyber defence tactics from scratch with multi-agent reinforcement learning, 2023. arXiv: 2310.05939."},{"key":"10.1016\/j.jisa.2026.104507_bib0025","series-title":"Artificial intelligence and machine learning for multi-domain operations applications III","first-page":"490","article-title":"Autonomous network cyber offence strategy through deep reinforcement learning","volume":"vol. 11746","author":"Sultana","year":"2021"},{"key":"10.1016\/j.jisa.2026.104507_bib0026","unstructured":"Andrew A., Spillard S., Collyer J., Dhir N.. Developing optimal causal cyber-defence agents via cyber security simulation, 2022. arXiv: 2207.12355."},{"key":"10.1016\/j.jisa.2026.104507_bib0027","series-title":"Artificial intelligence and machine learning for multi-domain operations applications V","first-page":"42","article-title":"Autonomous cyber warfare agents: dynamic reinforcement learning for defensive cyber operations","volume":"vol. 12538","author":"Bierbrauer","year":"2023"},{"key":"10.1016\/j.jisa.2026.104507_bib0028","series-title":"Proceedings of the AAAI\/ACM conference on AI, ethics, and society","first-page":"180","article-title":"Fooling LIME and SHAP: adversarial attacks on post hoc explanation methods","author":"Slack","year":"2020"},{"key":"10.1016\/j.jisa.2026.104507_bib0029","doi-asserted-by":"crossref","DOI":"10.1016\/j.patcog.2024.111221","article-title":"The level of strength of an explanation: a quantitative evaluation technique for post-hoc XAI methods","volume":"161","author":"Bello","year":"2025","journal-title":"Pattern Recognit"},{"key":"10.1016\/j.jisa.2026.104507_bib0030","series-title":"Proceedings of the AAAI conference on artificial intelligence","first-page":"2514","article-title":"Generation of policy-level explanations for reinforcement learning","volume":"vol. 33","author":"Topin","year":"2019"},{"key":"10.1016\/j.jisa.2026.104507_bib0031","first-page":"12222","article-title":"Edge: explaining deep reinforcement learning policies","volume":"34","author":"Guo","year":"2021","journal-title":"Adv Neural Inf Process Syst"},{"key":"10.1016\/j.jisa.2026.104507_bib0032","unstructured":"Deshmukh S.V., Dasgupta A., Krishnamurthy B., Jiang N., Agarwal C., Theocharous G., et al. Explaining RL decisions with trajectories, 2023. arXiv: 2305.04073."},{"key":"10.1016\/j.jisa.2026.104507_bib0033","series-title":"2024 IEEE Conference on artificial intelligence (CAI)","first-page":"75","article-title":"Abstracted trajectory visualization for explainability in reinforcement learning","author":"Takagi","year":"2024"},{"key":"10.1016\/j.jisa.2026.104507_bib0034","unstructured":"Li Y.. Deep reinforcement learning: an overview, 2017. arXiv: 1701.07274."},{"key":"10.1016\/j.jisa.2026.104507_bib0035","series-title":"Proceedings of the 22nd international conference on tools with artificial intelligence (ICTAI)","first-page":"243","article-title":"Adaptive \u03f5-greedy exploration in reinforcement learning","author":"Tokic","year":"2010"},{"issue":"3-4","key":"10.1016\/j.jisa.2026.104507_bib0036","first-page":"279","article-title":"Q-learning","volume":"8","author":"Watkins","year":"1992","journal-title":"Mach Learn"},{"key":"10.1016\/j.jisa.2026.104507_bib0037","series-title":"Dynamic programming","author":"Bellman","year":"1957"},{"key":"10.1016\/j.jisa.2026.104507_bib0038","series-title":"Proceedings of the 4th international conference on learning representations (ICLR)","article-title":"Prioritized experience replay","author":"Schaul","year":"2016"}],"container-title":["Journal of Information Security and Applications"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S2214212626001377?httpAccept=text\/xml","content-type":"text\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S2214212626001377?httpAccept=text\/plain","content-type":"text\/plain","content-version":"vor","intended-application":"text-mining"}],"deposited":{"date-parts":[[2026,7,15]],"date-time":"2026-07-15T15:47:03Z","timestamp":1784130423000},"score":1,"resource":{"primary":{"URL":"https:\/\/linkinghub.elsevier.com\/retrieve\/pii\/S2214212626001377"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9]]},"references-count":38,"alternative-id":["S2214212626001377"],"URL":"https:\/\/doi.org\/10.1016\/j.jisa.2026.104507","relation":{},"ISSN":["2214-2126"],"issn-type":[{"value":"2214-2126","type":"print"}],"subject":[],"published":{"date-parts":[[2026,9]]},"assertion":[{"value":"Elsevier","name":"publisher","label":"This article is maintained by"},{"value":"Unveiling the black box: A multi-layer framework for explaining reinforcement learning-based cyber agents","name":"articletitle","label":"Article Title"},{"value":"Journal of Information Security and Applications","name":"journaltitle","label":"Journal Title"},{"value":"https:\/\/doi.org\/10.1016\/j.jisa.2026.104507","name":"articlelink","label":"CrossRef DOI link to publisher maintained version"},{"value":"article","name":"content_type","label":"Content Type"},{"value":"\u00a9 2026 The Authors. Published by Elsevier Ltd.","name":"copyright","label":"Copyright"}],"article-number":"104507"}}