Relational Abstractions for Generalized Reinforcement Learning on Symbolic Problems
This work addresses the problem of long-horizon reasoning in symbolic reinforcement learning for AI researchers, offering a novel approach for efficient knowledge transfer without hand-coded curricula.
The paper tackles the challenge of reinforcement learning in symbolic state spaces by introducing a method that uses relational abstractions with deep learning to learn a generalizable Q-function, enabling zero-shot transfer to related problems with different object names and quantities, as shown in empirical evaluations on various problem instances.
Reinforcement learning in problems with symbolic state spaces is challenging due to the need for reasoning over long horizons. This paper presents a new approach that utilizes relational abstractions in conjunction with deep learning to learn a generalizable Q-function for such problems. The learned Q-function can be efficiently transferred to related problems that have different object names and object quantities, and thus, entirely different state spaces. We show that the learned generalized Q-function can be utilized for zero-shot transfer to related problems without an explicit, hand-coded curriculum. Empirical evaluations on a range of problems show that our method facilitates efficient zero-shot transfer of learned knowledge to much larger problem instances containing many objects.