Snake RL

Danger: Is there a wall or tail straight, left, or right?
Food Direction: Is food ahead, right, left, or behind (relative to head)?

Train a snake to eat food using Q-Learning.

Score: 0

High Score: 0

Episode: 0

Training Speed: 1x Exploration ($\epsilon$): 0.1

The agent doesn't see the whole grid. It only sees:

This simplifies the state space, allowing it to learn very quickly!