1 paper across 1 session
We investigate how digit bias learned from skewed pretraining data lead to systematic numerical hallucinations in LLMs.
