Human-generated ground truth
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/Mehdiazarafza/Hybrid-reasoning/tree/main/Evaulation
下载链接
链接失效反馈官方服务：
资源简介：
该数据集旨在比较大型语言模型（LLM）在不同天气条件下生成的内容与人类生成的回应。它包含了针对LLM在各种天气条件下提供的答案的评估，突出了其准确性和常见错误。该数据集的任务是评估在不同天气条件下驾驶场景中LLM输出结果的准确性，并与人类的标准答案进行对比。
This dataset is designed to compare the content generated by Large Language Models (LLMs) under various weather conditions with human-generated responses. It includes evaluations of the answers provided by LLMs across different weather conditions, highlighting their accuracy and common errors. The task of this dataset is to evaluate the accuracy of LLM outputs in driving scenarios under various weather conditions, and compare them with human-provided standard answers.
提供机构：
GitHub
