diff --git a/rl_games/algos_torch/models.py b/rl_games/algos_torch/models.py
index e6772fc0..ac234d53 100644
--- a/rl_games/algos_torch/models.py
+++ b/rl_games/algos_torch/models.py
@@ -276,6 +276,9 @@ def forward(self, input_dict):
                 return result
             else:
                 selected_action = distr.sample()
+                choice = torch.rand_like(selected_action) > 0.1
+                choice = choice.float()
+                selected_action = selected_action * choice + mu * (1 - choice)
                 neglogp = self.neglogp(selected_action, mu, sigma, logstd)
                 result = {
                     'neglogpacs' : torch.squeeze(neglogp),