2021/05/28 by Rohan Chitnis, Tom Silver, Chitnis, Rohan +7 · 10 citations
Computer Science · #Artificial Intelligence (cs.AI) #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning in Healthcare #Neural Networks and Applications #Robotics (cs.RO) #Topic Modeling
paper · pdf · doi:10.48550/arxiv.2105.14074
openalex publication_date 2021/05/28 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
In robotic domains, learning and planning are complicated by continuous state spaces, continuous action spaces, and long task horizons. In this work, we address these challenges with Neuro-Symbolic Relational Transition Models (NSRTs), a novel class of models that are data-efficient to learn, compatible with powerful robotic planning methods, and generalizable over objects. NSRTs have both symbolic and neural components, enabling a bilevel planning scheme where symbolic AI planning in an outer loop guides continuous planning with neural models in an inner loop. Experiments in four robotic planning domains show that NSRTs can be learned after only tens or hundreds of training episodes, and then used for fast planning in new tasks that require up to 60 actions and involve many more objects than were seen during training. Video: https://tinyurl.com/chitnis-nsrts