li-lab/tutorqa
收藏Hugging Face2025-04-27 更新2025-10-18 收录
下载链接：
https://hf-mirror.com/datasets/li-lab/tutorqa
下载链接
链接失效反馈官方服务：
资源简介：
TutorQA基准数据集是用于评估推理、图理解和语言生成各个方面性能的测试集，包含6个设计用于不同评估目的的任务。每个任务都是一个独立的split，包括关系判断、先决条件预测、路径搜索、子图补全、聚类和想法仓鼠（开放式问题，没有答案）。
The TutorQA Benchmark is a test set designed for evaluating performance on various aspects of reasoning, graph understanding, and language generation, containing 6 tasks tailored for different evaluation purposes. Each task is a separate split, including Relation Judgment, Prerequisite Prediction, Path Searching, Subgraph Completion, Clustering, and Idea Hamster (open-ended, no answers).
提供机构：
li-lab
