Advancing Large Language Model Reasoning Techniques: Methods Enabling LLMs to 'Think' Beyond Text Generation for Reliable and Explainable AI