Introduction to Reinforcement Learning and Its Role in LLMs