Optimal Skipping Rates: Training Agents with Fine-Grained Control Using Deep Reinforcement Learning

Figure 5

Showing the agent learning and testing performance for 3-11 optimal skip counts only.