fangrulin/redial
收藏Hugging Face2024-12-05 更新2024-12-14 收录
下载链接：
https://hf-mirror.com/datasets/fangrulin/redial
下载链接
链接失效反馈官方服务：
资源简介：
这是一个通过重写7个现有基准创建的逻辑推理数据集，重写工作由非裔美国人白话英语（AAVE）使用者完成，并经过人类和语言模型的质量检查。该数据集旨在用于研究测试语言模型对AAVE使用者的推理公平性和鲁棒性。
This is a reasoning dataset created by rewriting 7 existing benchmarks. The rewriting is conducted by human African American Vernacular English (AAVE) speakers, and quality check is conducted by both humans and language models. This dataset is used for research on testing language models reasoning fairness and robustness towards AAVE users.
提供机构：
fangrulin
