Generating new concepts with hybrid neuro-symbolic models

TL;DR

This paper introduces a hybrid neuro-symbolic model that combines neural networks and symbolic probabilistic programs to generate novel visual concepts, demonstrating improved generalization and representation over purely neural models.

Contribution

The paper presents a novel neuro-symbolic approach for concept generation that integrates structured symbolic knowledge with neural networks, advancing the synthesis of statistical and symbolic cognition.

Findings

Hybrid model outperforms neural-only models in likelihood on held-out classes.

Hybrid model produces more convincing and diverse generated concepts.

Hybrid approach generalizes better from limited training data.

Abstract

Human conceptual knowledge supports the ability to generate novel yet highly structured concepts, and the form of this conceptual knowledge is of great interest to cognitive scientists. One tradition has emphasized structured knowledge, viewing concepts as embedded in intuitive theories or organized in complex symbolic knowledge structures. A second tradition has emphasized statistical knowledge, viewing conceptual knowledge as an emerging from the rich correlational structure captured by training neural networks and other statistical models. In this paper, we explore a synthesis of these two traditions through a novel neuro-symbolic model for generating new concepts. Using simple visual concepts as a testbed, we bring together neural networks and symbolic probabilistic programs to learn a generative model of novel handwritten characters. Two alternative models are explored with more generic neural network architectures. We compare each of these three models for their likelihoods on held-out character classes and for the quality of their productions, finding that our hybrid model learns the most convincing representation and generalizes further from the training observations.