diff --git a/evoprompt/optimization.py b/evoprompt/optimization.py
index 29e711184e38497b30c576086f2d246f6e611d51..045d777cff0d4e93e6a69fbd25f26f43843d4ee8 100644
--- a/evoprompt/optimization.py
+++ b/evoprompt/optimization.py
@@ -167,7 +167,7 @@ class PromptOptimization:
         self.judge_model = judge_model
         self.run_options = run_options
 
-        self.judge_action = run_options.get("judge_action", "none")
+        self.judge_action = run_options.get("judge_action")
 
     def evaluate_prompt(self, prompt: str, parents: tuple[Prompt] | None = None):
         parent_histories = (