SaketR1/biasrm-annotations-paper-replication

该数据集包含四个字段：quaternion（四元数，可能表示旋转或方向），best和worst（可能表示某种评估或排名），以及annotator_llm（可能是标注者或模型的名称）。数据集分为训练集，包含8799个示例，总大小为3480092字节。没有提供详细的数据集用途或背景信息。

The dataset includes four fields: quaternion (possibly representing rotation or orientation), best and worst (possibly indicating some form of evaluation or ranking), and annotator_llm (might be the name of an annotator or a model). The dataset is split into a training set with 8799 examples, totaling 3480092 bytes in size. No detailed information about the datasets purpose or context is provided.