This research provides a comprehensive survey of reasoning strategies for Large Language Models (LLMs), addressing the critical gap between their language proficiency and actual reasoning capabilities.

• Reasoning Gap: Despite impressive language skills, LLMs struggle with complex reasoning tasks that require deeper cognitive processing
• Strategic Framework: The paper catalogs and analyzes multiple reasoning approaches to enhance LLM performance
• Current Limitations: Identifies where today's models fall short in reasoning capabilities
• Future Directions: Outlines promising paths for improvement in reasoning-capable AI

Medical Impact: As healthcare increasingly relies on AI systems, reasoning-capable LLMs could transform clinical decision support, diagnostic assistance, and treatment planning while maintaining appropriate safety guardrails.

Thinking Machines: A Survey of LLM based Reasoning Strategies