RicciNets: Curvature-guided Pruning of High-performance Neural Networks Using Ricci Flow
This work addresses the challenge of compressing high-performance neural networks for efficiency, though it is incremental as it builds on existing pruning methods with a new topological approach.
The authors tackled the problem of pruning randomly wired neural networks before training by using Ricci curvature to identify and remove low-importance edges, resulting in a 35% reduction in FLOPs with no performance degradation.
A novel method to identify salient computational paths within randomly wired neural networks before training is proposed. The computational graph is pruned based on a node mass probability function defined by local graph measures and weighted by hyperparameters produced by a reinforcement learning-based controller neural network. We use the definition of Ricci curvature to remove edges of low importance before mapping the computational graph to a neural network. We show a reduction of almost $35\%$ in the number of floating-point operations (FLOPs) per pass, with no degradation in performance. Further, our method can successfully regularize randomly wired neural networks based on purely structural properties, and also find that the favourable characteristics identified in one network generalise to other networks. The method produces networks with better performance under similar compression to those pruned by lowest-magnitude weights. To our best knowledge, this is the first work on pruning randomly wired neural networks, as well as the first to utilize the topological measure of Ricci curvature in the pruning mechanism.