The limitations of ROUGE lead to false negatives, where factually correct responses are marked as incorrect, and false positives, where incorrect responses receive high scores.

Authors

• Person: Not available Organization: arXiv
  Re-evaluating Hallucination Detection in LLMs - arXiv

Sources

• Re-evaluating Hallucination Detection in LLMs - arXiv arxiv.org arXiv via serper

Referenced by nodes (1)

• ROUGE concept