Tracking Uncensored General Intelligence in AI Models
-
-
-
Range: 0-100
Measures the model's knowledge breadth on uncensored topics. Higher scores indicate greater capability to understand and process diverse, potentially sensitive information.
Range: 0-10
Indicates the model's willingness to engage with controversial topics. Higher scores show less restrictive content filtering.
π‘ Note: The UGI Leaderboard uses undisclosed evaluation questions to prevent benchmark gaming and maintain fairness across models.
# | Model | UGI Score π Uncensored General Intelligence: Knowledge breadth on sensitive topics (0-100) | W/10 Score π Willingness to answer: Engagement with controversial topics (0-10) | Ideology |
---|---|---|---|---|
Loading leaderboard data... |