Back to Home

Reinforcement Learning Playground

Train AI agents to solve tasks using various RL algorithms

Grid World Environment

How it Works

The agent (blue) must navigate to the goal (green) while avoiding obstacles (red). The agent learns through trial and error, receiving rewards for reaching the goal and penalties for hitting obstacles. Watch as the agent gradually improves its strategy!

Control Panel
Q-Learning: Off-policy TD control using max Q-value for action selection
0
Episodes
0
Total Reward
0.0
Avg Reward
0%
Success Rate
Reward History
Q-Table Visualization