diff --git a/evaluate/evaluate_by_scoring.py b/evaluate/evaluate_by_scoring.py
index bf05185..301e9e6 100644
--- a/evaluate/evaluate_by_scoring.py
+++ b/evaluate/evaluate_by_scoring.py
@@ -23,6 +23,8 @@ def llm_model_func(prompt, system_prompt=None, history_messages=[], **kwargs) ->
     response = openai_client.chat.completions.create(
         model=LLM_MODEL, messages=messages, **kwargs
     )
+    if not response.choices or response.choices[0].message is None:
+        raise ValueError("LLM returned empty or filtered response")
     return response.choices[0].message.content
 
 
diff --git a/evaluate/evaluate_by_selection.py b/evaluate/evaluate_by_selection.py
index bfbccf5..4336edc 100644
--- a/evaluate/evaluate_by_selection.py
+++ b/evaluate/evaluate_by_selection.py
@@ -23,6 +23,8 @@ def llm_model_func(prompt, system_prompt=None, history_messages=[], **kwargs) ->
     response = openai_client.chat.completions.create(
         model=LLM_MODEL, messages=messages, **kwargs
     )
+    if not response.choices or response.choices[0].message is None:
+        raise ValueError("LLM returned empty or filtered response")
     return response.choices[0].message.content
 
 
diff --git a/reproduce/Step_2_extract_question.py b/reproduce/Step_2_extract_question.py
index e3960b1..5bed435 100644
--- a/reproduce/Step_2_extract_question.py
+++ b/reproduce/Step_2_extract_question.py
@@ -28,6 +28,8 @@ def llm_model_func(prompt, system_prompt=None, history_messages=[], **kwargs) ->
     response = openai_client.chat.completions.create(
         model=LLM_MODEL, messages=messages, **kwargs
     )
+    if not response.choices or response.choices[0].message is None:
+        raise ValueError("LLM returned empty or filtered response")
     return response.choices[0].message.content