基于神经符号程序合成与任务生成的抽象推理问题求解学习
Learning to Solve Abstract Reasoning Problems with Neurosymbolic Program Synthesis and Task Generation
AI总结：
本研究提出基于神经程序合成的抽象推理方法TransCoder，以类型化领域特定语言为核心，利用失败任务程序生成带解的合成数据监督训练，在抽象推理语料库数据集上生成数万合成问题，推动学习取得系统性进展。
摘要：
抽象思考与类比推理能力是快速适应新环境、通过分解处理新遇到的问题以及综合知识全面解决问题的前提。我们提出TransCoder——一种基于神经程序合成的抽象问题求解方法，并对所提架构中生成模块做出的决策开展了全面分析。TransCoder的核心是一种类型化领域特定语言，旨在为特征工程与抽象推理提供便利。训练阶段，我们利用未能求解任务的程序生成新任务，并将这些任务整合为一个合成数据集。由于以这种方式生成的每个合成任务都有已知的关联程序（即解），模型以监督模式在这些任务上进行训练。解以透明的程序形式呈现，可被检查与验证。我们在Abstract Reasoning Corpus（抽象推理语料库）数据集上验证了TransCoder的性能，我们的框架为该数据集生成了数万个带对应解的合成问题，助力学习过程取得系统性进展。
英文摘要：
The ability to think abstractly and reason by analogy is a prerequisite to rapidly adapt to new conditions, tackle newly encountered problems by decomposing them, and synthesize knowledge to solve problems comprehensively. We present TransCoder, a method for solving abstract problems based on neural program synthesis, and conduct a comprehensive analysis of decisions made by the generative module of the proposed architecture. At the core of TransCoder is a typed domain-specific language, designed to facilitate feature engineering and abstract reasoning. In training, we use the programs that failed to solve tasks to generate new tasks and gather them in a synthetic dataset. As each synthetic task created in this way has a known associated program (solution), the model is trained on them in supervised mode. Solutions are represented in a transparent programmatic form, which can be inspected and verified. We demonstrate TransCoder's performance using the Abstract Reasoning Corpus dataset, for which our framework generates tens of thousands of synthetic problems with corresponding solutions and facilitates systematic progress in learning.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
