Reinforcement Learning and Data-Generation for Syntax-Guided Synthesis

被引:0
|
作者
Parsert, Julian [1 ,2 ]
Polgreen, Elizabeth [2 ]
机构
[1] Univ Oxford, Oxford, England
[2] Univ Edinburgh, Edinburgh, Midlothian, Scotland
基金
奥地利科学基金会; 英国工程与自然科学研究理事会;
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
Program synthesis is the task of automatically generating code based on a specification. In Syntax-Guided Synthesis (SyGuS) this specification is a combination of a syntactic template and a logical formula, and the result is guaranteed to satisfy both. We present a reinforcement-learning guided algorithm for SyGuS which uses Monte-Carlo Tree Search (MCTS) to search the space of candidate solutions. Our algorithm learns policy and value functions which, combined with the upper confidence bound for trees, allow it to balance exploration and exploitation. A common challenge in applying machine learning approaches to syntax-guided synthesis is the scarcity of training data. To address this, we present a method for automatically generating training data for SyGuS based on anti-unification of existing first-order satisfiability problems, which we use to train our MCTS policy. We implement and evaluate this setup and demonstrate that learned policy and value improve the synthesis performance over a baseline by over 26 percentage points in the training and testing sets. Our tool outperforms state-of-the-art tool cvc5 on the training set and performs comparably in terms of the total number of problems solved on the testing set (solving 23% of the benchmarks on which cvc5 fails). We make our data set publicly available, to enable further application of machine learning methods to the SyGuS problem.
引用
收藏
页码:10670 / 10678
页数:9
相关论文
共 50 条
  • [41] SG-Net: Syntax-Guided Machine Reading Comprehension
    Zhang, Zhuosheng
    Wu, Yuwei
    Zhou, Junru
    Duan, Sufeng
    Zhao, Hai
    Wang, Rui
    THIRTY-FOURTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THE THIRTY-SECOND INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE AND THE TENTH AAAI SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2020, 34 : 9636 - 9643
  • [42] Syntax-Guided Content-Adaptive Transform for Image Compression
    Shi, Yunhui
    Ye, Liping
    Wang, Jin
    Wang, Lilong
    Hu, Hui
    Yin, Baocai
    Ling, Nam
    SENSORS, 2024, 24 (16)
  • [43] NL2Viz: Natural Language to Visualization via Constrained Syntax-Guided Synthesis
    Wu, Zhengkai
    Vu Le
    Tiwari, Ashish
    Gulwani, Sumit
    Radhakrishna, Arjun
    Radicek, Ivan
    Soares, Gustavo
    Wang, Xinyu
    Li, Zhenwen
    Xie, Tao
    PROCEEDINGS OF THE 30TH ACM JOINT MEETING EUROPEAN SOFTWARE ENGINEERING CONFERENCE AND SYMPOSIUM ON THE FOUNDATIONS OF SOFTWARE ENGINEERING, ESEC/FSE 2022, 2022, : 972 - 983
  • [44] Syntax-guided Neural Module Distillation to Probe Compositionality in Sentence Embeddings
    Pandey, Rohan
    17TH CONFERENCE OF THE EUROPEAN CHAPTER OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, EACL 2023, 2023, : 3143 - 3149
  • [45] Data Boost: Text Data Augmentation Through Reinforcement Learning Guided Conditional Generation
    Liu, Ruibo
    Xu, Guangxuan
    Jia, Chenyan
    Ma, Weicheng
    Wang, Lili
    Vosoughi, Soroush
    PROCEEDINGS OF THE 2020 CONFERENCE ON EMPIRICAL METHODS IN NATURAL LANGUAGE PROCESSING (EMNLP), 2020, : 9031 - 9041
  • [46] ROLE-PLAY AS A DATA-GENERATION METHOD
    ROSENDALE, D
    SIMULATION & GAMING, 1989, 20 (04) : 487 - 492
  • [47] Extracting relational facts based on hybrid Syntax-Guided transformer and pointer network
    Liu, Luping
    Wang, Meiling
    He, Xiaohai
    Qing, Linbo
    Zhang, Jin
    JOURNAL OF INTELLIGENT & FUZZY SYSTEMS, 2021, 40 (06) : 12167 - 12183
  • [48] Synthetic Data: Legal Implications of the Data-Generation Revolution
    Gal, Michal S.
    Lynskey, Orla
    IOWA LAW REVIEW, 2024, 109 (03) : 1087 - 1156
  • [49] Dependency to Semantics: Structure Transformation and Syntax-Guided Attention for Neural Semantic Parsing
    Wu, Shan
    Chen, Bo
    Han, Xianpei
    Sun, Le
    KNOWLEDGE GRAPH AND SEMANTIC COMPUTING: KNOWLEDGE GRAPH EMPOWERS NEW INFRASTRUCTURE CONSTRUCTION, 2021, 1466 : 119 - 133
  • [50] Optimal Data-Generation Strategy for Machine Learning Yield Functions in Anisotropic Plasticity
    Shoghi, Ronak
    Hartmaier, Alexander
    FRONTIERS IN MATERIALS, 2022, 9