Episode
0
Step
0
Reward
0
Episode len: 0
Avg reward(100): 0
0.00
Reinforce policy gradient demo — 4 inputs (x, x_dot, theta, theta_dot) → 6 hidden → 3 outputs (left, none, right).