diff --git a/README.md b/README.md
index 208edbe..e9f88a1 100644
--- a/README.md
+++ b/README.md
@@ -31,7 +31,7 @@ Currently we have the following environements:
 |`Rosenbrock{dim}-v0`| Gym interface for Rosenbrock function. `{dim}` is one of 5, 10, 25, 50 or 100. | 
 
 
-## INSTALL
+## Install
 1. Clone the repository 
 ```bash 
 git clone git@github.com:ALRhub/alr_envs.git
diff --git a/alr_envs/__init__.py b/alr_envs/__init__.py
index 2193db9..62e5db0 100644
--- a/alr_envs/__init__.py
+++ b/alr_envs/__init__.py
@@ -64,7 +64,7 @@ register(
 )
 
 register(
-    id='ALRReacherSparse-v0',
+    id='ALRReacher7Sparse-v0',
     entry_point='alr_envs.mujoco:ALRReacherEnv',
     max_episode_steps=200,
     kwargs={
diff --git a/alr_envs/mujoco/alr_reacher.py b/alr_envs/mujoco/alr_reacher.py
index 7ae28da..0c65e7f 100644
--- a/alr_envs/mujoco/alr_reacher.py
+++ b/alr_envs/mujoco/alr_reacher.py
@@ -37,13 +37,16 @@ class ALRReacherEnv(mujoco_env.MujocoEnv, utils.EzPickle):
 
         reward_dist = 0.0
         angular_vel = 0.0
+        reward_balance = 0.0
         if self._steps >= self.steps_before_reward:
             vec = self.get_body_com("fingertip") - self.get_body_com("target")
             reward_dist -= self.reward_weight * np.linalg.norm(vec)
             angular_vel -= np.linalg.norm(self.sim.data.qvel.flat[:self.n_links])
         reward_ctrl = - np.square(a).sum()
-        reward_balance = - self.balance_weight * np.abs(
-            angle_normalize(np.sum(self.sim.data.qpos.flat[:self.n_links]), type="rad"))
+
+        if self.balance:
+            reward_balance = - self.balance_weight * np.abs(
+                angle_normalize(np.sum(self.sim.data.qpos.flat[:self.n_links]), type="rad"))
 
         reward = reward_dist + reward_ctrl + angular_vel + reward_balance
         self.do_simulation(a, self.frame_skip)