Object-Oriented Transition Modeling with Inductive Logic Programming

ICLR 2026 Conference Submission667 Authors

Keywords: reinforcement learning, transition modeling, inductive logic program, decision trees

TL;DR: We describe a novel algorithm for learning transition and reward models in the framework of object-oriented reinforcement learning.

Abstract: Building models of the world from observation, i.e., *induction*, is one of the major challenges in machine learning. In order to be useful, models need to maintain accuracy when used in novel situations, i.e., generalize. In addition, they should be easy to interpret and efficient to train. Prior work has investigated these concepts in the context of *object-oriented representations* inspired by human cognition. In this paper, we develop a novel learning algorithm that is substantially more powerful than these previous methods. Our thorough experiments, including ablation tests and comparison with neural baselines, demonstrate a significant improvement over the state-of-the-art. The source code for all of our algorithms and benchmarks will be available online after publication.

Primary Area: reinforcement learning

Submission Number: 667