diff --git a/acme/agents/jax/mpo/config.py b/acme/agents/jax/mpo/config.py
index 3e1d70ab31..596cf29b3b 100644
--- a/acme/agents/jax/mpo/config.py
+++ b/acme/agents/jax/mpo/config.py
@@ -32,8 +32,9 @@ class MPOConfig:
   discrete_policy: bool = False
 
   # Specification of the type of experience the learner will consume.
-  experience_type: mpo_types.ExperienceType = mpo_types.FromTransitions(
-      n_step=5)
+  experience_type: mpo_types.ExperienceType = dataclasses.field(
+      default_factory=lambda: mpo_types.FromTransitions(n_step=5)
+  )
   num_stacked_observations: int = 1
   # Optional data-augmentation transformation for observations.
   observation_transform: Optional[Callable[[types.NestedTensor],