The Scattering Compositional Learner: Discovering Objects, Attributes, Relationships in Analogical Reasoning
This addresses the problem of compositional reasoning in AI for tasks like analogical reasoning, with incremental improvements in performance and robustness.
The paper tackles analogical reasoning on Raven's Progressive Matrices by proposing the Scattering Compositional Learner (SCL), which achieves state-of-the-art performance with a 48.7% relative improvement on Balanced-RAVEN and 26.4% on PGM, and discovers compositional representations of objects and relationships.
In this work, we focus on an analogical reasoning task that contains rich compositional structures, Raven's Progressive Matrices (RPM). To discover compositional structures of the data, we propose the Scattering Compositional Learner (SCL), an architecture that composes neural networks in a sequence. Our SCL achieves state-of-the-art performance on two RPM datasets, with a 48.7% relative improvement on Balanced-RAVEN and 26.4% on PGM over the previous state-of-the-art. We additionally show that our model discovers compositional representations of objects' attributes (e.g., shape color, size), and their relationships (e.g., progression, union). We also find that the compositional representation makes the SCL significantly more robust to test-time domain shifts and greatly improves zero-shot generalization to previously unseen analogies.