Unused Exploration Bonus

ml/env.py

@@ -78,6 +78,15 @@ class JackBotEnv(gym.Env):
         self.target_height = 0.14
         self.collapse_height_fraction = 0.55
         self.tilt_failure_rad = 0.9
+
+        # Small random exploration pulse to break local optima.
+        # The bonus is only granted on a rare step and only when the robot
+        # is still stable enough that a bold move is likely to remain safe.
+        self.exploration_bonus_prob = 0.03
+        self.exploration_bonus_interval = 120
+        self.exploration_bonus_scale = 0.08
+        self.exploration_bonus_active = False
+
         self.start_positions = [[0.0, 0.0, 0.0]]
         self.max_distance_from_start = [0.0]
         self.max_survival_steps = 0
@@ -157,6 +166,7 @@ class JackBotEnv(gym.Env):
         self.last_action = np.zeros(self.action_space.shape[0], dtype=np.float32)
         self.max_distance_from_start = [0.0]
         self.max_survival_steps = 0
+        self.exploration_bonus_active = False
 
         if self._first_reset:
             self.curriculum_phase = 0