Bálint K.

The training approach uses hardware to demonstrate actions—motors physically pull the trigger and drive the ball to show the rat what to do before they try it themselves. It's reinforcement learning implemented in hardware rather than pure trial-and-error. What you get: - All 3D-printable designs, circuit diagrams, firmware, and control software on GitHub. Build guide included. v1 was featured on Vice and PC Gamer. v2 is fully modular and ready to replicate.