AI RESEARCH
Answer-Set-Programming-based Abstractions for Reinforcement Learning
arXiv CS.AI
•
ArXi:2605.31444v1 Announce Type: new Reinforcement Learning (RL) enables autonomous agents to learn policies from experience, but realistic problems often involve enormous state spaces, making learning and generalisation challenging. Abstraction and approximation are therefore essential. Relational Reinforcement Learning (RRL) offers a way to reason about objects and their relations, and the CARCASS framework by Martijn van Otterlo nstrates how logical representations can model Marko Decision Processes (MDPs) in first-order domains.