Performance-Aware Reinforcement Learning Framework for LLM-Augmented Decision Agents in Complex Environments
International Journal of Information and Communication Technology Trends · 01.04.2026 · 6 (1) · 57-63
DOI: 10.66525/ijictt74 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Information and Communication Technology Trends, online erschienen am 01.04.2026. Verfasst von Oliver Smith und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
This study presents a performance-aware reinforcement learning framework for enhancing decision-making efficiency in LLM-augmented agent systems. The framework incorporates multi-objective reward functions that jointly consider task accuracy, computational cost and response latency. A large-scale experimental evaluation is conducted on 12,600 decision tasks, including financial analysis, operational planning and scenario simulation. The proposed method achieves a 20.9% increase in decision accuracy while reducing average computation time by 16.7%. Additionally, Pareto frontier analysis shows improved trade-offs between performance and efficiency compared to baseline approaches. The results indicate that performance-aware optimization can significantly enhance the practical applicability of intelligent agent systems.
Zitieren
APA
Smith, O., Thompson, J., & Brown, W. (2026). Performance-Aware Reinforcement Learning Framework for LLM-Augmented Decision Agents in Complex Environments. International Journal of Information and Communication Technology Trends, 6(1), 57-63. https://doi.org/10.66525/ijictt74
Vancouver
Smith O, Thompson J, Brown W. Performance-Aware Reinforcement Learning Framework for LLM-Augmented Decision Agents in Complex Environments. International Journal of Information and Communication Technology Trends. 2026;6(1):57-63. doi:10.66525/ijictt74
BibTeX
@article{Smith2026,
author = {Smith, Oliver and Thompson, James and Brown, William},
title = {Performance-Aware Reinforcement Learning Framework for LLM-Augmented Decision Agents in Complex Environments},
journal = {International Journal of Information and Communication Technology Trends},
year = {2026},
volume = {6},
number = {1},
pages = {57--63},
doi = {10.66525/ijictt74}
}
EUVIMED