Learning to Prove Theorems by Learning to Generate Theorems

被引:0
|
作者
Wang, Mingzhe [1 ]
Deng, Jia [1 ]
机构
[1] Princeton Univ, Princeton, NJ 08544 USA
来源
ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 33, NEURIPS 2020 | 2020年 / 33卷
基金
美国国家科学基金会;
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
We consider the task of automated theorem proving, a key AI task. Deep learning has shown promise for training theorem provers, but there are limited human-written theorems and proofs available for supervised learning. To address this limitation, we propose to learn a neural generator that automatically synthesizes theorems and proofs for the purpose of training a theorem prover. Experiments on real-world tasks demonstrate that synthetic data from our approach improves the theorem prover and advances the state of the art of automated theorem proving in Metamath. Code is available at https://github.com/princeton-vl/MetaGen.
引用
收藏
页数:12
相关论文
共 50 条
  • [31] Improving the time efficiency of proving theorems using a learning mechanism
    Almonayyes, A
    Raafat, H
    Almulla, M
    Alharshani, R
    INTERNATIONAL JOURNAL OF COMPUTER MATHEMATICS, 2001, 77 (02) : 169 - 180
  • [32] Proving Theorems using Incremental Learning and Hindsight Experience Replay
    Aygun, Eser
    Anand, Ankit
    Orseau, Laurent
    Glorot, Xavier
    McAleer, Stephen
    Firoiu, Vlad
    Zhang, Lei
    Precup, Doina
    Mourad, Shibl
    INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 162, 2022,
  • [33] Regularized learning in Banach spaces as an optimization problem: representer theorems
    Zhang, Haizhang
    Zhang, Jun
    JOURNAL OF GLOBAL OPTIMIZATION, 2012, 54 (02) : 235 - 250
  • [34] Regularized learning in Banach spaces as an optimization problem: representer theorems
    Haizhang Zhang
    Jun Zhang
    Journal of Global Optimization, 2012, 54 : 235 - 250
  • [36] Sparse Representer Theorems for Learning in Reproducing Kernel Banach Spaces
    Wang, Rui
    Xu, Yuesheng
    Yan, Mingsong
    JOURNAL OF MACHINE LEARNING RESEARCH, 2024, 25
  • [37] Learning to Generate Questions by Learning What not to Generate
    Liu, Bang
    Zhao, Mingjun
    Niu, Di
    Lai, Kunfeng
    He, Yancheng
    Wei, Haojie
    Xu, Yu
    WEB CONFERENCE 2019: PROCEEDINGS OF THE WORLD WIDE WEB CONFERENCE (WWW 2019), 2019, : 1106 - 1118
  • [38] A Method to Prove Japanese Theorems and Others Appeared in Wasan Using Maxima
    Takato, Setsuo
    Makishita, Hideyo
    SYMBOLIC COMPUTATION IN SOFTWARE SCIENCE, SCSS 2024, 2024, 14991 : 57 - 78
  • [39] A Method to Prove Japanese Theorems and Others Appeared in Wasan Using Maxima
    Takato, Setsuo
    Makishita, Hideyo
    Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2024, 14991 LNAI : 57 - 78