Hybrid LLM-Augmented Reinforcement Learning Agents for Complex Sequential Decision Tasks cs.AI · 2026-08-04