遇见数据集

yale-nlp/SciReas-Pro

收藏
Hugging Face2025-09-18 更新2025-10-18 收录
官方服务:

资源简介:

SciReas-Pro是一个推理密集型的科学问题解决数据集,用于评估大型语言模型在科学问题解决方面的知识和推理能力。该数据集由10个现有基准组成,是从SciReas数据集中选取的子集,包含了问题、答案、任务名称、文档ID等信息,并针对每个问题提供了是否正确的标记以及是否为推理密集型任务的标记。

SciReas-Pro is a reasoning-intensive scientific problem-solving dataset designed to evaluate the knowledge and reasoning capabilities of large language models in scientific problem-solving. This dataset consists of 10 existing benchmarks and is a subset selected from the SciReas dataset, including information such as questions, answers, task names, and document IDs, and provides correctness labels and reasoning-intensive task labels for each question.

提供机构:
yale-nlp
二维码
社区交流群
二维码
科研交流群
商业服务