Unsupervised Learning of Neurosymbolic Encoders

TL;DR

This paper introduces an unsupervised framework for learning neurosymbolic encoders that combine neural networks with symbolic programs, improving interpretability and category separation in real-world trajectory data.

Contribution

The authors develop a novel method integrating program synthesis with VAEs to learn neurosymbolic encoders that incorporate expert knowledge for better interpretability.

Findings

Better separation of meaningful categories than standard VAEs

Practical improvements in behavior classification tasks

Enhanced interpretability of latent representations

Abstract

We present a framework for the unsupervised learning of neurosymbolic encoders, which are encoders obtained by composing neural networks with symbolic programs from a domain-specific language. Our framework naturally incorporates symbolic expert knowledge into the learning process, which leads to more interpretable and factorized latent representations compared to fully neural encoders. We integrate modern program synthesis techniques with the variational autoencoding (VAE) framework, in order to learn a neurosymbolic encoder in conjunction with a standard decoder. The programmatic descriptions from our encoders can benefit many analysis workflows, such as in behavior modeling where interpreting agent actions and movements is important. We evaluate our method on learning latent representations for real-world trajectory data from animal biology and sports analytics. We show that our approach offers significantly better separation of meaningful categories than standard VAEs and leads to practical gains on downstream analysis tasks, such as for behavior classification.