Table 1 Assessment of the quality of questions and answers generated by Answer-first QA generation.
From: A hallucination detection and mitigation framework for faithful text summarization using LLMs
ROUGE | Recall | Precision | F1 |
|---|---|---|---|
ROUGE-1 | 51.02 | 78.13 | 61.73 |
ROUGE-2 | 44.78 | 71.42 | 55.04 |
ROUGE-L | 51.02 | 78.12 | 61.72 |
