Do Language Models Know When They’re Hallucinating References? F i n d i n g s - E A C L