CVMar 26, 2024

Towards Explaining Hypercomplex Neural Networks

arXiv:2403.17929v16 citationsh-index: 26Has CodeIJCNN
Originality Incremental advance
AI Analysis

This work addresses the problem of explaining hypercomplex neural networks for researchers and practitioners in deep learning, representing an incremental advancement in interpretability methods for this specific domain.

The paper tackles the lack of interpretability in hypercomplex neural networks by proposing inherently interpretable parameterized hypercomplex neural networks (PHNNs) and quaternion-like networks, achieving this through a cosine-similarity transform that reduces the model to a single linear transform for direct interpretation.

Hypercomplex neural networks are gaining increasing interest in the deep learning community. The attention directed towards hypercomplex models originates from several aspects, spanning from purely theoretical and mathematical characteristics to the practical advantage of lightweight models over conventional networks, and their unique properties to capture both global and local relations. In particular, a branch of these architectures, parameterized hypercomplex neural networks (PHNNs), has also gained popularity due to their versatility across a multitude of application domains. Nonetheless, only few attempts have been made to explain or interpret their intricacies. In this paper, we propose inherently interpretable PHNNs and quaternion-like networks, thus without the need for any post-hoc method. To achieve this, we define a type of cosine-similarity transform within the parameterized hypercomplex domain. This PHB-cos transform induces weight alignment with relevant input features and allows to reduce the model into a single linear transform, rendering it directly interpretable. In this work, we start to draw insights into how this unique branch of neural models operates. We observe that hypercomplex networks exhibit a tendency to concentrate on the shape around the main object of interest, in addition to the shape of the object itself. We provide a thorough analysis, studying single neurons of different layers and comparing them against how real-valued networks learn. The code of the paper is available at https://github.com/ispamm/HxAI.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes