A Study of Compositional Generalization in Neural Models

Understand

Compositional and relational learning is a hallmark of human intelligence, but one which presents challenges for neural models.

• One difficulty in the development of such models is the lack of benchmarks with clear compositional and relational task structure on which to systematically evaluate them.
• In this paper, we introduce an environment called ConceptWorld, which enables the generation of images from compositional and relational concepts, defined using a logical domain specific language.
• We use it to generate images for a variety of compositional structures: 2x2 squares, pentominoes, sequences, scenes involving these objects, and other more complex concepts.