# Evolving Connectivity for Recurrent Spiking Neural Networks

arXiv: https://arxiv.org/abs/2305.17650 · HTML: https://mlsanity.com/paper/2305.17650/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 28 May 2023.

## Built on by (most cited first, 6 of 6)

- [Edge Intelligence with Spiking Neural Networks](https://mlsanity.com/paper/2507.14069/) (2025)
- [Adaptive Surrogate Gradients for Sequential Reinforcement Learning in Spiking Neural Networks](https://mlsanity.com/paper/2510.24461/) (2025)
- [Sleep-Based Homeostatic Regularization for Stabilizing Spike-Timing-Dependent Plasticity in Recurrent Spiking Neural Networks](https://mlsanity.com/paper/2601.08447/) (2026)
- [Signal-Adaptive Trust Regions for Gradient-Free Optimization of Recurrent Spiking Neural Networks](https://mlsanity.com/paper/2601.21572/) (2026)
- [Learning fast changing slow in spiking neural networks](https://mlsanity.com/paper/2402.10069/) (2024)
- [NeuroTrain: Surveying Local Learning Rules for Spiking Neural Networks with an Open Benchmarking Framework](https://mlsanity.com/paper/2605.15058/) (2026)

## Abstract (arXiv)

Recurrent spiking neural networks (RSNNs) hold great potential for advancing artificial general intelligence, as they draw inspiration from the biological nervous system and show promise in modeling complex dynamics. However, the widely-used surrogate gradient-based training methods for RSNNs are inherently inaccurate and unfriendly to neuromorphic hardware. To address these limitations, we propose the evolving connectivity (EC) framework, an inference-only method for training RSNNs. The EC framework reformulates weight-tuning as a search into parameterized connection probability distributions, and employs Natural Evolution Strategies (NES) for optimizing these distributions. Our EC framework circumvents the need for gradients and features hardware-friendly characteristics, including sparse boolean connections and high scalability. We evaluate EC on a series of standard robotic locomotion tasks, where it achieves comparable performance with deep neural networks and outperforms gradient-trained RSNNs, even solving the complex 17-DoF humanoid task. Additionally, the EC framework demonstrates a two to three fold speedup in efficiency compared to directly evolving parameters. By providing a performant and hardware-friendly alternative, the EC framework lays the groundwork for further energy-efficient applications of RSNNs and advances the development of neuromorphic devices.