{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,25]],"date-time":"2026-08-25T22:26:51Z","timestamp":1787696811004,"version":"build-2784847793"},"reference-count":30,"publisher":"Elsevier BV","license":[{"start":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T00:00:00Z","timestamp":1777593600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/tdm\/userlicense\/1.0\/"},{"start":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T00:00:00Z","timestamp":1777593600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/legal\/tdmrep-license"},{"start":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T00:00:00Z","timestamp":1777593600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-017"},{"start":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T00:00:00Z","timestamp":1777593600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"},{"start":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T00:00:00Z","timestamp":1777593600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-012"},{"start":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T00:00:00Z","timestamp":1777593600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T00:00:00Z","timestamp":1777593600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-004"}],"content-domain":{"domain":["elsevier.com","sciencedirect.com"],"crossmark-restriction":true},"short-container-title":["Neurocomputing"],"published-print":{"date-parts":[[2026,5]]},"DOI":"10.1016\/j.neucom.2026.133080","type":"journal-article","created":{"date-parts":[[2026,2,17]],"date-time":"2026-02-17T20:05:23Z","timestamp":1771358723000},"page":"133080","update-policy":"https:\/\/doi.org\/10.1016\/elsevier_cm_policy","source":"Crossref","is-referenced-by-count":1,"special_numbering":"C","title":["Interpreting multi-agent reinforcement learning decisions via key feature activation"],"prefix":"10.1016","volume":"677","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-9538-1472","authenticated-orcid":false,"given":"Peizhang","family":"Li","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1276-8511","authenticated-orcid":false,"given":"Qing","family":"Fei","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zhen","family":"Chen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zhongqi","family":"Sun","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bo","family":"Wang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","reference":[{"key":"10.1016\/j.neucom.2026.133080_bib0005","author":"Ackermann"},{"key":"10.1016\/j.neucom.2026.133080_bib0010","series-title":"Advances in Neural Information Processing Systems","first-page":"4699","article-title":"Neural additive models: interpretable machine learning with neural nets","author":"Agarwal","year":"2021"},{"key":"10.1016\/j.neucom.2026.133080_bib0015","first-page":"4561","article-title":"Towards better interpretability in deep Q-Networks","volume":"33","author":"Annasamy","year":"2019","journal-title":"Proc. AAAI Conf. Artif. Intell."},{"key":"10.1016\/j.neucom.2026.133080_bib0020","doi-asserted-by":"crossref","first-page":"16","DOI":"10.1016\/j.eng.2021.10.007","article-title":"From unmanned systems to autonomous intelligent systems","volume":"12","author":"Chen","year":"2022","journal-title":"Engineering"},{"key":"10.1016\/j.neucom.2026.133080_bib0025","doi-asserted-by":"crossref","first-page":"1226","DOI":"10.1109\/COMST.2021.3063822","article-title":"Single and Multi-Agent deep reinforcement learning for AI-enabled wireless networks: a tutorial","volume":"23","author":"Feriani","year":"2021","journal-title":"IEEE Commun. Surv. Tutorials"},{"key":"10.1016\/j.neucom.2026.133080_bib0030","doi-asserted-by":"crossref","DOI":"10.1016\/j.neucom.2024.128482","article-title":"An improved hierarchical deep reinforcement learning algorithm for multi-intelligent vehicle lane change","volume":"609","author":"Gao","year":"2024","journal-title":"Neurocomputing"},{"key":"10.1016\/j.neucom.2026.133080_bib0035","series-title":"International Conference on Machine Learning","first-page":"2961","article-title":"Actor-attention-critic for multi-agent reinforcement learning","author":"Iqbal","year":"2019"},{"key":"10.1016\/j.neucom.2026.133080_bib0040","doi-asserted-by":"crossref","DOI":"10.1016\/j.neucom.2024.127818","article-title":"Data-driven adaptive consensus control for heterogeneous nonlinear Multi-Agent systems using online reinforcement learning","volume":"596","author":"Ji","year":"2024","journal-title":"Neurocomputing"},{"key":"10.1016\/j.neucom.2026.133080_bib0045","series-title":"IJCAI\/ECAI Workshop on Explainable Artificial Intelligence","article-title":"Explainable reinforcement learning via reward decomposition","author":"Juozapaitis","year":"2019"},{"key":"10.1016\/j.neucom.2026.133080_bib0050","author":"Klein"},{"key":"10.1016\/j.neucom.2026.133080_bib0055","series-title":"Advances in Neural Information Processing Systems","article-title":"Multi-agent actor-critic for mixed cooperative-competitive environments","author":"Lowe","year":"2017"},{"key":"10.1016\/j.neucom.2026.133080_bib0060","series-title":"Advances in Neural Information Processing Systems","article-title":"A unified approach to interpreting model predictions","volume":"vol. 30","author":"Lundberg","year":"2017"},{"key":"10.1016\/j.neucom.2026.133080_bib0065","doi-asserted-by":"crossref","DOI":"10.1016\/j.neucom.2024.128111","article-title":"Explainable artificial intelligence: a survey of needs, techniques, applications, and future direction","volume":"599","author":"Mersha","year":"2024","journal-title":"Neurocomputing"},{"key":"10.1016\/j.neucom.2026.133080_bib0070","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3616864","article-title":"Explainable reinforcement learning: a survey and comparative review","volume":"56","author":"Milani","year":"2024","journal-title":"ACM Comput. Surv."},{"key":"10.1016\/j.neucom.2026.133080_bib0075","series-title":"Proceedings of the Conference on Fairness, Accountability, and Transparency","first-page":"279","article-title":"Explaining explanations in AI","author":"Mittelstadt","year":"2019"},{"key":"10.1016\/j.neucom.2026.133080_bib0080","doi-asserted-by":"crossref","first-page":"529","DOI":"10.1038\/nature14236","article-title":"Human-level control through deep reinforcement learning","volume":"518","author":"Mnih","year":"2015","journal-title":"Nature"},{"key":"10.1016\/j.neucom.2026.133080_bib0085","doi-asserted-by":"crossref","first-page":"1371","DOI":"10.1109\/TAI.2022.3222143","article-title":"Hierarchical reinforcement learning for AIR combat at DARPA\u2019s AlphaDogfight trials","volume":"4","author":"Pope","year":"2023","journal-title":"IEEE Trans. Artif. Intell."},{"key":"10.1016\/j.neucom.2026.133080_bib0090","series-title":"Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining","first-page":"1135","article-title":"\u201cwhy should i trust you?\u201d: explaining the predictions of any classifier","author":"Ribeiro","year":"2016"},{"key":"10.1016\/j.neucom.2026.133080_bib0095","doi-asserted-by":"crossref","first-page":"16693","DOI":"10.1007\/s00521-022-07280-8","article-title":"Hierarchical goals contextualize local reward decomposition explanations","volume":"35","author":"Rietz","year":"2023","journal-title":"Neural Comput. Appl."},{"key":"10.1016\/j.neucom.2026.133080_bib0100","series-title":"Advances in Neural Information Processing Systems","article-title":"Dynamic routing between capsules","author":"Sabour","year":"2017"},{"key":"10.1016\/j.neucom.2026.133080_bib0105","doi-asserted-by":"crossref","first-page":"4531","DOI":"10.1109\/TNSM.2021.3096673","article-title":"Multi-Agent DRL for task offloading and resource allocation in multi-UAV enabled IoT edge network","volume":"18","author":"Seid","year":"2021","journal-title":"IEEE Trans. Netw. Serv. Manag."},{"key":"10.1016\/j.neucom.2026.133080_bib0110","series-title":"Proceedings of the IEEE International Conference on Computer Vision","first-page":"618","article-title":"Grad-CAM: visual explanations from deep networks via gradient-based localization","author":"Selvaraju","year":"2017"},{"key":"10.1016\/j.neucom.2026.133080_bib0115","doi-asserted-by":"crossref","DOI":"10.1016\/j.artint.2020.103367","article-title":"Interestingness elements for explainable reinforcement learning: understanding agents\u2019 capabilities and limitations","volume":"288","author":"Sequeira","year":"2020","journal-title":"Artif. Intell."},{"key":"10.1016\/j.neucom.2026.133080_bib0120","doi-asserted-by":"crossref","first-page":"10222","DOI":"10.1109\/TPAMI.2021.3133717","article-title":"Temporal-spatial causal interpretations for vision-based reinforcement learning","volume":"44","author":"Shi","year":"2021","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"10.1016\/j.neucom.2026.133080_bib0125","series-title":"Proceedings of the 31st International Conference on Machine Learning","first-page":"387","article-title":"Deterministic policy gradient algorithms","author":"Silver","year":"2014"},{"key":"10.1016\/j.neucom.2026.133080_bib0130","series-title":"Artificial Intelligence and Machine Learning","first-page":"69","article-title":"Task independent Capsule-Based agents for deep Q-Learning","author":"Singh","year":"2022"},{"key":"10.1016\/j.neucom.2026.133080_bib0135","series-title":"Proceedings of the 35th International Conference on Machine Learning","first-page":"5045","article-title":"Programmatically interpretable reinforcement learning","author":"Verma","year":"2018"},{"key":"10.1016\/j.neucom.2026.133080_bib0140","doi-asserted-by":"crossref","DOI":"10.1016\/j.neucom.2024.127904","article-title":"A deep reinforcement learning based distributed multi-UAV dynamic area coverage algorithm for complex environment","volume":"595","author":"Xiao","year":"2024","journal-title":"Neurocomputing"},{"key":"10.1016\/j.neucom.2026.133080_bib0145","first-page":":5:1","article-title":"Reinforcement learning in healthcare: a survey","volume":"55","author":"Yu","year":"2021","journal-title":"ACM Comput. Surv."},{"key":"10.1016\/j.neucom.2026.133080_bib0150","first-page":"24611","article-title":"The surprising effectiveness of PPO in cooperative Multi-Agent games","volume":"35","author":"Yu","year":"2022","journal-title":"Adv. Neural Inf. Process. Syst."}],"container-title":["Neurocomputing"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S0925231226004777?httpAccept=text\/xml","content-type":"text\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S0925231226004777?httpAccept=text\/plain","content-type":"text\/plain","content-version":"vor","intended-application":"text-mining"}],"deposited":{"date-parts":[[2026,4,1]],"date-time":"2026-04-01T06:24:12Z","timestamp":1775024652000},"score":1,"resource":{"primary":{"URL":"https:\/\/linkinghub.elsevier.com\/retrieve\/pii\/S0925231226004777"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5]]},"references-count":30,"alternative-id":["S0925231226004777"],"URL":"https:\/\/doi.org\/10.1016\/j.neucom.2026.133080","relation":{},"ISSN":["0925-2312"],"issn-type":[{"value":"0925-2312","type":"print"}],"subject":[],"published":{"date-parts":[[2026,5]]},"assertion":[{"value":"Elsevier","name":"publisher","label":"This article is maintained by"},{"value":"Interpreting multi-agent reinforcement learning decisions via key feature activation","name":"articletitle","label":"Article Title"},{"value":"Neurocomputing","name":"journaltitle","label":"Journal Title"},{"value":"https:\/\/doi.org\/10.1016\/j.neucom.2026.133080","name":"articlelink","label":"CrossRef DOI link to publisher maintained version"},{"value":"article","name":"content_type","label":"Content Type"},{"value":"\u00a9 2026 Elsevier B.V. All rights are reserved, including those for text and data mining, AI training, and similar technologies.","name":"copyright","label":"Copyright"}],"article-number":"133080"}}