Update app.py

app.py

@@ -77,15 +77,16 @@ async def run_and_submit_all(profile: gr.OAuthProfile | None):
                 "You are a precise and concise AI assistant competing in a strict evaluation benchmark.  You will be asked a question. You must only output a single final answer, using the following format exactly: FINAL ANSWER: [your answer]"
                 "Guidelines:
                 Do NOT explain your reasoning.
-                Do NOT output markdown, sections, titles, bullet points, or formatting like "Task outcome", "Detailed version", or "Additional context".
+                Do NOT output markdown, sections, titles, bullet points, or formatting like (Task outcome), (Detailed version), or (Additional context).
                 Do NOT include references, sources, or citations.
-                Do NOT include any leading text like "Here is the final answer", "The correct answer is", or "Answer:".
+                Do NOT include any leading text like (Here is the final answer), (The correct answer is), or (Answer):.
                 Do NOT repeat the question or summarize it.
                 Your response must consist of a single line starting with: FINAL ANSWER: and nothing else.
                 Normalization rules: 
                 If the answer is a list, apply the above rules to each item, and separate them with commas.
                 If the answer is a number, write it as digits (no commas, no units unless asked).
-                If the answer is a string, use lowercase, no articles ("a", "an", "the"), and no abbreviations."
+                If the answer is a string, use lowercase, no articles (a, an, the), and no abbreviations."
+                "If there is no correct answer (e.g., no such person, place, or event), respond with: FINAL ANSWER: none"
                 "If you are asked for a comma separated list, apply the above rules depending of whether the element to be put in the list is a number or a string.\n\n"
             
             )