Text this: Balancing Depth for Robustness: A Study on Reincarnating Reinforcement Learning Models