diff --git a/source/unitree_rl_lab/unitree_rl_lab/tasks/locomotion/agents/rsl_rl_ppo_cfg.py b/source/unitree_rl_lab/unitree_rl_lab/tasks/locomotion/agents/rsl_rl_ppo_cfg.py
index f5f98fbf..26317632 100644
--- a/source/unitree_rl_lab/unitree_rl_lab/tasks/locomotion/agents/rsl_rl_ppo_cfg.py
+++ b/source/unitree_rl_lab/unitree_rl_lab/tasks/locomotion/agents/rsl_rl_ppo_cfg.py
@@ -14,8 +14,11 @@ class BasePPORunnerCfg(RslRlOnPolicyRunnerCfg):
     save_interval = 100
     experiment_name = ""  # same as task name
     empirical_normalization = False
+    # Clip actions passed to the environment to prevent action-rate/reward explosions
+    clip_actions = 1.0
     policy = RslRlPpoActorCriticCfg(
         init_noise_std=1.0,
+        noise_std_type="log",
         actor_hidden_dims=[512, 256, 128],
         critic_hidden_dims=[512, 256, 128],
         activation="elu",