Leaderboard
percent correct, lowest to highest
AGI Detector
…

is AGI here yet?
yet another countless benchmark - stupid, specific and unapologetically difficult.
evaluates how well leading AI models know Cat Bot. web search is disabled, meaning models must inherently know Cat Bot and have it in their training data to score anything above zero.
runs just 5 questions, each more niche than the last!
percent correct, lowest to highest