diff --git a/rllib/models/tf/tf_action_dist.py b/rllib/models/tf/tf_action_dist.py
index 27813ddd3..06c182372 100644
--- a/rllib/models/tf/tf_action_dist.py
+++ b/rllib/models/tf/tf_action_dist.py
@@ -335,6 +335,14 @@ class SquashedGaussian(TFActionDistribution):
             axis=-1)
         return log_prob
 
+    @override(ActionDistribution)
+    def entropy(self) -> TensorType:
+        raise ValueError("Entropy not defined for SquashedGaussian!")
+
+    @override(ActionDistribution)
+    def kl(self, other: ActionDistribution) -> TensorType:
+        raise ValueError("KL not defined for SquashedGaussian!")
+
     def _squash(self, raw_values: TensorType) -> TensorType:
         # Returned values are within [low, high] (including `low` and `high`).
         squashed = ((tf.math.tanh(raw_values) + 1.0) / 2.0) * \
diff --git a/rllib/models/torch/torch_action_dist.py b/rllib/models/torch/torch_action_dist.py
index 83f900a5e..ecc8aa276 100644
--- a/rllib/models/torch/torch_action_dist.py
+++ b/rllib/models/torch/torch_action_dist.py
@@ -243,6 +243,14 @@ class TorchSquashedGaussian(TorchDistributionWrapper):
             torch.log(1 - unsquashed_values_tanhd**2 + SMALL_NUMBER), dim=-1)
         return log_prob
 
+    @override(TorchDistributionWrapper)
+    def entropy(self) -> TensorType:
+        raise ValueError("Entropy not defined for SquashedGaussian!")
+
+    @override(TorchDistributionWrapper)
+    def kl(self, other: ActionDistribution) -> TensorType:
+        raise ValueError("KL not defined for SquashedGaussian!")
+
     def _squash(self, raw_values: TensorType) -> TensorType:
         # Returned values are within [low, high] (including `low` and `high`).
         squashed = ((torch.tanh(raw_values) + 1.0) / 2.0) * \