1 paper · 1 filter
Yang Liu, Bin Chong, Yuhan Lin +7
Large language models (LLMs) frequently produce \emph{detail hallucinations} when processing long regulatory documents, including subtle errors in threshold values, units, scopes,…