rowshq/aispreadsheetbenchmark
收藏Hugging Face2025-09-22 更新2025-11-01 收录
下载链接:
https://hf-mirror.com/datasets/rowshq/aispreadsheetbenchmark
下载链接
链接失效反馈官方服务:
资源简介:
AI Spreadsheet Benchmark数据集是一个包含53个现实电子表格提示的基准,用于评估电子表格辅助工具在现实情境下的表现,如是否编写公式、图表是否与数据链接、当数字变化时输出是否可以重新计算等。该数据集分为经典数据分析、高级分析、创建模型、管理电子表格元素和算术运算五大类别,每个提示记录包含自然语言指令和类别及子类别标签。
The AI Spreadsheet Benchmark dataset consists of 53 realistic spreadsheet prompts designed to evaluate the behavior of spreadsheet copilots in situ, including whether they write formulas, whether charts stay linked to data, and whether the output can recompute when numbers change. The dataset is categorized into five main areas: Classic Data Analysis, Advanced Analysis, Creating Models, Managing Spreadsheet Elements, and Arithmetic Operations, with each prompt record containing natural-language instructions and category and sub-category labels.
提供机构:
rowshq



