Adaptive Network Reconfiguration for Limiting Memory Poisoning Spread in Multi-Agent Collaboration
Global Media and Social Sciences Research Journal · 25.03.2026 · 7 (1) · 272-279
DOI: 10.71465/gmssrj173 ↗Kurz gefasst
Fachartikel in der Zeitschrift Global Media and Social Sciences Research Journal, online erschienen am 25.03.2026. Verfasst von Lars de Vries und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).
Abstract (Originalsprache)
Propagation of memory poisoning is influenced by the topology of agent communication networks. This study proposes an adaptive network reconfiguration strategy that dynamically rewires communication edges based on anomaly scores. Agents compute local contamination likelihood and temporarily suspend high-risk communication links. A topology optimization algorithm minimizes global contamination probability while preserving connectivity. Simulations were conducted on 200-agent networks with varying degree distributions. Under a 25% poisoning rate, adaptive reconfiguration reduced contamination coverage from 71.2% to 28.9% and improved task success rate by 38.4%. Network connectivity remained above 0.92 during reconfiguration. Dynamic topology control effectively constrains poisoning spread in collaborative systems.
Zitieren
APA
Vries, L. D., Dijk, S. V., & Bakker, J. (2026). Adaptive Network Reconfiguration for Limiting Memory Poisoning Spread in Multi-Agent Collaboration. Global Media and Social Sciences Research Journal, 7(1), 272-279. https://doi.org/10.71465/gmssrj173
Vancouver
Vries LD, Dijk SV, Bakker J. Adaptive Network Reconfiguration for Limiting Memory Poisoning Spread in Multi-Agent Collaboration. Global Media and Social Sciences Research Journal. 2026;7(1):272-279. doi:10.71465/gmssrj173
BibTeX
@article{Vries2026,
author = {Vries, Lars de and Dijk, Sophie van and Bakker, Jeroen},
title = {Adaptive Network Reconfiguration for Limiting Memory Poisoning Spread in Multi-Agent Collaboration},
journal = {Global Media and Social Sciences Research Journal},
year = {2026},
volume = {7},
number = {1},
pages = {272--279},
doi = {10.71465/gmssrj173}
}
EUVIMED