Reinforcement Learning
Navigate three mazes yourself, then train a Q-learning robot. Watch its learned policy and compare the routes it finds.
0/3 discoveries
Find the safe route
Move the robot from the top left to the flag without touching a trap.
Take the model apart.
Click a piece. See its job in slow motion.
Preparing your experiment
Loading the model and its controls…
Made for curious minds.
Explore reinforcement learning with an interactive model running in your browser.
Find your next experiment →