Toward Compositional Generalization in Object-Oriented World Modeling

2022/04/28 by Zhao, Linfeng, Kong, Lingzhi, Walters, Robin +1 · 3 citations
#Artificial Intelligence (cs.AI) #FOS: Computer and information sciences #Machine Learning (cs.LG) #Robotics (cs.RO)

paper · doi:10.48550/arxiv.2204.13661

Abstract

Compositional generalization is a critical ability in learning and decision-making. We focus on the setting of reinforcement learning in object-oriented environments to study compositional generalization in world modeling. We (1) formalize the compositional generalization problem with an algebraic approach and (2) study how a world model can achieve that. We introduce a conceptual environment, Object Library, and two instances, and deploy a principled pipeline to measure the generalization ability. Motivated by the formulation, we analyze several methods with exact or no compositional generalization ability using our framework, and design a differentiable approach, Homomorphic Object-oriented World Model (HOWM), that achieves soft but more efficient compositional generalization.

Cited by

• From monoliths to modules: Decomposing transducers for efficient world modelling
• Learning Interactive World Model for Object-Centric Reinforcement Learning
• What Drives Compositional Generalization in Visual Generative Models?

Related

• Concepts, Properties and an Approach for Compositional Generalization
• Successes and Limitations of Object-centric Models at Compositional Generalisation
• Compositional Generalization from First Principles
• Learning Dynamic Attribute-factored World Models for Efficient Multi-object Reinforcement Learning
• Generalization properties of contrastive world models