MPTP2078 Benchmark
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/BartoszPiotrowski/stateful-premise-selection-with-RNNs
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2078个用于训练机器学习模型的小米扎定理，旨在为定理提出有用的前提。数据集涵盖了定理和前提的各种表示方法，同时还包括了数据增强的方法。规模上，该数据集拥有1469个定理，包含24087个独特的证明和2227个不同的前提。所承担的任务是为定理证明进行前提选择。
This dataset contains 2078 Xiao Miza theorems designed for training machine learning models, with the objective of providing useful premises for the corresponding theorems. It covers diverse representation methods for both theorems and their premises, and also includes data augmentation techniques. In terms of scale, this dataset comprises 1469 theorems, 24087 unique proofs, and 2227 distinct premises. The core task supported by this dataset is premise selection for automated theorem proving.
提供机构：
Bartosz Piotrowski
