Search for a command to run...
Least-to-Most Prompting Enables Complex Reasoning in Large Language Models