Overconfident Assertion
LLMs present information assertively even when their internal uncertainty is high, favoring strong declarative statements over probabilistic ones.
Examples (1)
This method is the most effective approach for solving this class of problems.
Strength: Likely AI71%
-100%100%