Learning to Learn Generative Programs with Memoised Wake-Sleep.
Conference on Uncertainty in Artificial Intelligence(2020)
摘要
We study a class of neuro-symbolic generative models in which neural networks are used both for inference and as priors over symbolic, data-generating programs. As generative models, these programs capture compositional structures in a naturally explainable form. To tackle the challenge of performing program induction as an 'inner-loop' to learning, we propose the Memoised Wake-Sleep (MWS) algorithm, which extends Wake Sleep by explicitly storing and reusing the best programs discovered by the inference network throughout training. We use MWS to learn accurate, explainable models in three challenging domains: strokebased character modelling, cellular automata, and few-shot learning in a novel dataset of real-world string concepts.
更多查看译文
关键词
Semi-Supervised Learning,Sequence-to-Sequence Learning
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要