Large Language Models (LLMs) are prone to generating hallucinated content, which compromises their reliability in knowledge-intensive tasks. To address this …
机构:华为
来源:arXiv 2610.03063 | AI4Papers 论文推荐平台