diff --git a/agent/asyncrl/policy_output.py b/agent/asyncrl/policy_output.py
index c7d55d3..001d192 100644
--- a/agent/asyncrl/policy_output.py
+++ b/agent/asyncrl/policy_output.py
@@ -18,11 +18,15 @@ def _sample_discrete_actions(batch_probs):
       List consisting of sampled actions
     """
     action_indices = []
-
-    # Subtract a tiny value from probabilities in order to avoid
-    # "ValueError: sum(pvals[:-1]) > 1.0" in numpy.multinomial
-    batch_probs = batch_probs - np.finfo(np.float32).epsneg
-
+    # Prevent having a vector which sum is not in [0, 1]
+    while not 0 < np.sum(batch_probs) < 1: 
+        # Subtract a tiny value from probabilities in order to avoid
+        # "ValueError: sum(pvals[:-1]) > 1.0" in numpy.multinomial
+        batch_probs = batch_probs - np.finfo(np.float32).epsneg
+        # Apply abs function to keep probability values positive to avoid
+        # "ValueError: pvals < 0, pvals > 1 or pvals contains NaNs" in numpy.multinomial
+        batch_probs = np.absolute(batch_probs)
+    
     for i in range(batch_probs.shape[0]):
         histogram = np.random.multinomial(1, batch_probs[i])
         action_indices.append(int(np.nonzero(histogram)[0]))