2020/07/08 by Yuhuai Wu, Honghua Dong, Wu, Yuhuai +5 · 26 citations
Computer Science · Mathematics · Psychology · #Algorithm #Analogical reasoning #Analogy #Artificial Intelligence (cs.AI) #Artificial intelligence #Cognition #Computer science #Domain (mathematical analysis) #Domain Adaptation and Few-Shot Learning #FOS: Computer and information sciences #Focus (optics) #Generalization #Logic in Computer Science (cs.LO) #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Mathematics #Natural Language Processing Techniques #Natural language processing #Pattern recognition (psychology) #Psychology #Raven's Progressive Matrices #Representation (politics) #State (computer science) #Task (project management) #Theoretical computer science #Topic Modeling #cs.AI #cs.LG #cs.LO #stat.ML
paper · pdf · doi:10.48550/arxiv.2007.04212
published in arXiv (Cornell University) (Cornell University)
arxiv created 2020/07/08 · openalex publication_date 2020/07/08 · arxiv updated 2020/07/09 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
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.