Box Embeddings: An open-source library for representation learning using geometric structures

Abstract

A major factor contributing to the success of modern representation learning is the ease of performing various vector operations. Recently, objects with geometric structures (eg. distributions, complex or hyperbolic vectors, or regions such as cones, disks, or boxes) have been explored for their alternative inductive biases and additional representational capacities. In this work, we introduce Box Embeddings, a Python library that enables researchers to easily apply and extend probabilistic box embeddings. 1 Fundamental geometric operations on boxes are implemented in a numerically stable way, as are modern approaches to training boxes which mitigate gradient sparsity. The library is fully open-source, and compatible with both PyTorch and TensorFlow, which allows existing neural network layers to be replaced with or transformed into boxes effortlessly. In this work, we present the implementation details of the fundamental components of the library, and the concepts required to use box representations alongside existing neural network architectures.

Event-Event Relation Extraction using Probabilistic Box Embedding

“…Recent works have successfully used box representations in conjunction with neural networks to represent input text for tasks like entity typing (Onoe et al, 2021), multi-label classification (Patel et al, 2022), natural language entailment (Chheda et al, 2021), etc. In all these works, the input is rep-…”