Tag
LongNovel introduces a multi-scale benchmark for evaluating hallucinations in long-context novel summarization, built from Chinese and English novels to improve LLM assessment.