Search for a command to run...
Benefits and Pitfalls of Reinforcement Learning for Language Model Planning: A Theoretical Perspective