Structured Reward functions using Signal Temporal Logic specifications

This package provides a few simple experiments for Bounded Horizon Nominal Robustness-based reward functions. The working examples are currently only CartPole.