Quantization Meets Spikes: Nearly Lossless Conversion at the First Timestep via Polarity Multi-Spike Mapping

2025/08/20 by Hangming Zhang, Zheng Li, Zhang, Hangming +7
Engineering · #FOS: Computer and information sciences #Neural and Evolutionary Computing (cs.NE) #Optical Polarization and Ellipsometry

paper · pdf · doi:10.48550/arxiv.2508.14520

openalex publication_date 2025/08/20 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Spiking neural networks (SNNs) offer advantages in computational efficiency via event-driven computing, compared to traditional artificial neural networks (ANNs). While direct training methods tackle the challenge of non-differentiable activation mechanisms in SNNs, they often suffer from high computational and energy costs during training. As a result, ANN-to-SNN conversion approach remains a valuable and practical alternative. These conversion-based methods aim to leverage the discrete output produced by the quantization layer to obtain SNNs with low latency. Although the theoretical minimum latency is one timestep, existing conversion methods have struggled to realize such ultra-low latency without accuracy loss. Moreover, current quantization approaches often discard negative-value information following batch normalization and are highly sensitive to the hyperparameter configuration, leading to degraded performance. In this work, we, for the first time, analyze the information loss introduced by quantization layers through the lens of information entropy. Building on our analysis, we introduce polarity multi-spike mapping (PMSM) framework and a hyperparameter initialization strategy tailored for the quantization layer. Our method achieves nearly lossless ANN-to-SNN conversion at the extremity, i.e., the first timestep, while also leveraging the temporal dynamics of SNNs across multiple timesteps to maintain stable performance on complex tasks. Extensive experiments on six image and neuromorphic datasets consistently demonstrate that PMSM achieves nearly lossless accuracy at the first timestep. Remarkably, despite operating under ultra-low-latency constraints, PMSM surpasses state-of-the-art direct training methods on multiple benchmarks.

Citations

• Efficient ANN-SNN Conversion with Error Compensation Learning
• Scaling Spike-Driven Transformer With Efficient Spike Firing Approximation Training
• Spiking Transformer with Spatial-Temporal Attention
• Converting High-Performance and Low-Latency SNNs through Explicit Modelling of Residual Error in ANNs
• QKFormer: Hierarchical Spiking Transformer using Q-K Attention
• SpikingResformer: Bridging ResNet and Vision Transformer in Spiking Neural Networks
• LM-HT SNN: Enhancing the Performance of SNN to ANN Counterpart through Learnable Multi-hierarchical Threshold Model
• Ternary Spike: Learning Ternary Spikes for Spiking Neural Networks
• RMP-Loss: Regularizing Membrane Potential Distribution for Spiking Neural Networks
• Spike-driven Transformer
• Fast-SNN: Fast Spiking Neural Network by Converting Quantized ANN
• A Hybrid Neural Coding Approach for Pattern Recognition with Spiking Neural Networks
• Bridging the Gap between ANNs and SNNs by Calibrating Offset Spikes
• Reducing ANN-SNN Conversion Error through Residual Membrane Potential
• Low Latency Conversion of Artificial Neural Network Models to Rate-encoded Spiking Neural Networks
• Training Spiking Neural Networks with Local Tandem Learning
• Masked Spiking Transformer
• Spikformer: When Spiking Neural Network Meets Transformer
• Attention Spiking Neural Networks
• Recurrent Bilinear Optimization for Binary Neural Networks
• Converting Artificial Neural Networks to Spiking Neural Networks via Parameter Calibration
• Optimized Potential Initialization for Low-latency Spiking Neural Networks
• Efficient Neuromorphic Signal Processing with Loihi 2
• A Free Lunch From ANN: Towards Efficient, Accurate Spiking Neural Networks Calibration
• Optimal ANN-SNN Conversion for Fast and Accurate Inference in Deep Spiking Neural Networks
• Optimal Conversion of Conventional Artificial Neural Networks to Spiking Neural Networks
• Going Deeper With Directly-Trained Larger Spiking Neural Networks
• An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
• TactileSGNet: A Spiking Graph Neural Network for Event-based Tactile Object Recognition
• RMP-SNN: Residual Membrane Potential Neuron for Enabling Deeper High-Accuracy and Low-Latency Spiking Neural Network
• Surrogate Gradient Learning in Spiking Neural Networks
• Surrogate Gradient Learning in Spiking Neural Networks: Bringing the Power of Gradient-Based Optimization to Spiking Neural Networks
• TactileGCN: A Graph Convolutional Network for Predicting Grasp Stability with Tactile Sensors
• SLAYER: Spike Layer Error Reassignment in Time
• Inductive Representation Learning on Large Graphs
• Deep Residual Learning for Image Recognition
• Very Deep Convolutional Networks for Large-Scale Image Recognition

Related

• Fast-SNN: Fast Spiking Neural Network by Converting Quantized ANN
• MSAT: Biologically Inspired Multi-Stage Adaptive Threshold for Conversion of Spiking Neural Networks
• One-Spike SNN: Single-Spike Phase Coding with Base Manipulation for ANN-to-SNN Conversion Loss Minimization
• Bridging the Gap between ANNs and SNNs by Calibrating Offset Spikes
• When Bio-Inspired Computing meets Deep Learning: Low-Latency, Accurate, & Energy-Efficient Spiking Neural Networks from Artificial Neural Networks