Intelligent-Internet/GAIA-Subset-Benchmark
收藏Hugging Face2025-07-01 更新2025-04-12 收录
下载链接：
https://hf-mirror.com/datasets/Intelligent-Internet/GAIA-Subset-Benchmark
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含了问题、多种布尔类型特征以及不同模型的报告。具体特征包括问题文本、是否困惑（Perplexity）、是否理解（Grok3）等，以及针对不同模型（如GPT-4o、Deepseek-R1、QwQ等）的报告文本。数据集分为benchmark部分，共包含44个示例。
The dataset includes questions, multiple boolean features, and reports from different models. Specific features include question text, whether it is confused (Perplexity), whether it is understood (Grok3), etc., as well as report text for different models (such as GPT-4o, Deepseek-R1, QwQ, etc.). The dataset is divided into a benchmark part, containing a total of 44 examples.
提供机构：
Intelligent-Internet
