SAA-Lab/LitBench-Test-Release

LitBench-Test-Release是一个经过优化的、完整的评论ID数据集，包含了所有必要列，不含任何元数据，专为数据集重放水工作流程设计。数据集由原始的LitBench-Test-IDs数据集增强而来，通过智能文本匹配找到了缺失的评论ID，并经过严格验证。最终包含2381行数据，每行都有两个Reddit评论ID，分别代表首选故事和较不首选故事的评论。

The LitBench-Test-Release is an optimized, complete comment ID dataset that contains only the essential columns, free of any metadata clutter, specifically designed for dataset rehydration workflows. The dataset is enhanced from the original LitBench-Test-IDs dataset by intelligent text matching to find missing comment IDs, followed by strict validation. It finally includes 2,381 rows with two Reddit comment IDs in each row, representing the preferred and less preferred story comments respectively.