2026. Retrieval-Augmented Reinforcement Learning for Multi-Hop Reasoning in Small Language Models. Journal of Data Intelligence and AI Systems. 1, 3 (Sep. 2026).