1 paper
James Xu Zhao, Jimmy Z. J. Liu, Bryan Hooi +1
Large language models (LLMs) are widely used for long-form text generation. However, factual errors in the responses would undermine their reliability. Despite growing attention to…