The Scattering Compositional Learner: Discovering Objects, Attributes, Relationships in Analogical Reasoning

Abstract

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.

References in corpus (8)

• A simple neural network module for relational reasoning
• Measuring abstract reasoning in neural networks
• Measuring Compositional Generalization: A Comprehensive Method on Realistic Data
• Automatically Composing Representation Transformations as a Means for Generalization
• Learning Perceptual Inference by Contrasting
• Differentiable Programs with Neural Libraries
• Abstract Reasoning with Distracting Features
• Stratified Rule-Aware Network for Abstract Visual Reasoning

Cited by in corpus (12)

• Abstraction and Analogy-Making in Artificial Intelligence
• Multi-Label Contrastive Learning for Abstract Visual Reasoning
• A Review of Emerging Research Directions in Abstract Visual Reasoning
• Emergent Symbols through Binding in External Memory
• Deep Learning Methods for Abstract Visual Reasoning: A Survey on Raven's Progressive Matrices
• Learning Visual Abstract Reasoning through Dual-Stream Networks
• Dynamic Inference with Neural Interpreters
• One Self-Configurable Model to Solve Many Abstract Visual Reasoning Problems
• Pairwise Relations Discriminator for Unsupervised Raven's Progressive Matrices
• Learning Algebraic Representation for Systematic Generalization in Abstract Reasoning
• Abstract Reasoning via Logic-guided Generation
• A Memory-Augmented Neural Network Model of Abstract Rule Learning