# BiSNN: Training Spiking Neural Networks with Binary Weights via Bayesian Learning

arXiv: https://arxiv.org/abs/2012.08300 · HTML: https://mlsanity.com/paper/2012.08300/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 15 Dec 2020.

## Built on by (most cited first, 8 of 8)

- [Bayesian Continual Learning via Spiking Neural Networks](https://mlsanity.com/paper/2208.13723/) (2022)
- [Knowing When to Stop: Delay-Adaptive Spiking Neural Network Classifiers with Reliability Guarantees](https://mlsanity.com/paper/2305.11322/) (2023)
- [Towards Accurate Binary Spiking Neural Networks: Learning with Adaptive Gradient Modulation Mechanism](https://mlsanity.com/paper/2502.14344/) (2025)
- [Bayesian Inference on Binary Spiking Networks Leveraging Nanoscale Device Stochasticity](https://mlsanity.com/paper/2302.01302/) (2023)
- [Twin Network Augmentation: A Novel Training Strategy for Improved Spiking Neural Networks and Efficient Weight Quantization](https://mlsanity.com/paper/2409.15849/) (2024)
- [Binary Spiking Neural Networks as Causal Models](https://mlsanity.com/paper/2604.27007/) (2026)
- [Bayes2IMC: In-Memory Computing for Bayesian Binary Neural Networks](https://mlsanity.com/paper/2411.07902/) (2024)
- [Training of Spiking Neural Networks with Expectation-Propagation](https://mlsanity.com/paper/2506.23757/) (2025)

## Abstract (arXiv)

Artificial Neural Network (ANN)-based inference on battery-powered devices can be made more energy-efficient by restricting the synaptic weights to be binary, hence eliminating the need to perform multiplications. An alternative, emerging, approach relies on the use of Spiking Neural Networks (SNNs), biologically inspired, dynamic, event-driven models that enhance energy efficiency via the use of binary, sparse, activations. In this paper, an SNN model is introduced that combines the benefits of temporally sparse binary activations and of binary weights. Two learning rules are derived, the first based on the combination of straight-through and surrogate gradient techniques, and the second based on a Bayesian paradigm. Experiments validate the performance loss with respect to full-precision implementations, and demonstrate the advantage of the Bayesian paradigm in terms of accuracy and calibration.