Instantiated Datasets for Deductive and Abductive Reasoning

该数据集包含了一系列实例化的逻辑问题，旨在用于评估演绎和归纳推理能力，分为12个不同的类别或领域，并设有4个难度级别。该数据集的目标是衡量大型语言模型在不同语境下的推理能力及其泛化潜力。此外，该数据集也可用于逻辑推理评估和模型微调任务。

This dataset comprises a series of instantiated logical problems, designed to evaluate deductive and inductive reasoning capabilities. It is categorized into 12 distinct categories or domains, with four difficulty levels established. The core objective of this dataset is to measure the reasoning abilities and generalization potential of large language models (LLMs) across diverse contexts. Additionally, this dataset can also be utilized for logical reasoning evaluation and model fine-tuning tasks.