Update app.py

app.py

@@ -218,9 +218,9 @@ with GraInter:
             <h2 style="margin-bottom: 0; font-size: 1.8em;">About</h2>
             <strong>UGI:</strong> Uncensored General Intelligence. A measurement of the amount of uncensored/controversial information an LLM knows and is willing to tell the user. It is calculated from the average score of 5 subjects LLMs commonly refuse to talk about. The leaderboard is made of roughly 65 questions/tasks, measuring both willingness to answer and accuracy in fact-based controversial questions. I'm choosing to keep the questions private so people can't train on them and devalue the leaderboard.
             
-            **W/10:** Willingness/10. A more narrow subset of the UGI questions, creating a 10-point score which measures how far the model can be pushed before going against its instructions, refusing to answer, or adding an ethical disclaimer to its response.
+            **W/10:** Willingness/10. A more narrow subset of the UGI questions, solely focused on measuring how far a model can be pushed before going against its instructions, refusing to answer, or adding an ethical disclaimer to its response.
             <br>
-            **I/10:** Intelligence/10. A 10-point score made up of the UGI questions with the highest correlation with parameter size. This metric shows how much a model's knowledge and reasoning play a role in its UGI score.
+            **I/10:** Intelligence/10. The average score of the UGI questions with the highest correlation with parameter size. This metric shows how much a model's knowledge and reasoning play a role in its UGI score.
             <br><br>
             A high UGI but low W/10 could mean for example that the model can provide a lot of accurate sensitive information, but will refuse to form the information into something it sees as dangerous. Or that it answers questions correctly, but appends a paragraph to its answer explaining why the question is immoral to ask.
             <br><br>