M. Dredze

LLM agents can achieve high task accuracy while exhibiting profound epistemic *im*humility, failing to acknowledge uncertainty even when presented with contradictory evidence.

Weird generalization is not just a quirk of model training; it’s a fragile phenomenon that could be weaponized if not carefully managed.

MLLMs struggle with visually rendered text not because they can't reason, but because they can't *read* it, and a simple self-distillation fix closes the gap.