提出可同时量化神经元膜电位的低比特SNN部署方法
PTQ4SNN: Membrane-Aware Post-Training Quantization for Spiking Neural Networks
- 通过通道级统一缩放桥约束膜电位尺度，适配分布差异
- 按放电活跃度与敏感度分配2/4/8比特精度，平均4比特下保持准确率
- 无需重训练，支持卷积SNN与脉冲驱动Transformer
脉冲神经网络（SNN）实现稀疏事件驱动计算，但其低比特部署仍不完善，因循环膜电位常以浮点保留，即使权重已量化。量化这些电位困难在于其分布跨通道差异大，且接近阈值的小扰动可能改变脉冲决策并随时间累积。我们提出PTQ4SNN，一种仅需小校准集的膜电位感知后训练量化框架，联合量化权重与循环膜电位。首先，通道级统一缩放桥将膜电位尺度设为s_mem,c = s_w,c * 2^k_c，适应膜电位分布并支持移位兼容缩放转换。其次，混合精度比特分配根据放电活跃度与量化敏感度，为膜通道分配2/4/8比特，在平均4比特预算下实现。该框架基于可复用投影-LIF对，无需主干网络重训练，适用于卷积SNN与脉冲驱动Transformer。静态与事件驱动分类及语义分割实验表明，PTQ4SNN在W4量化和约4比特膜电位精度下有效保持模型准确率。
原文摘要 · Abstract (English)
Spiking neural networks (SNNs) enable sparse and event-driven computation, but their low-bit deployment remains incomplete because recurrent membrane states are commonly retained in floating point even after weight quantization. Quantizing these states is challenging because their distributions differ across channels and from the preceding weights, while small perturbations near the firing threshold may alter spike decisions and accumulate over time. We propose PTQ4SNN, a membrane-aware post-training quantization framework that jointly quantizes weights and recurrent membrane states using only a small calibration set. First, a channel-wise Unified Scale Bridge constrains the membrane scale as s_mem,c = s_w,c * 2^k_c, adapting to membrane distributions while enabling shift-compatible scale conversion. Second, Mixed-Precision Bit Allocation assigns 2/4/8-bit precision to membrane channels according to firing activity and quantization sensitivity under an average-bit budget. The framework operates on reusable projection-LIF pairs and supports both convolutional SNNs and spike-driven Transformers without backbone retraining. Experiments on static and event-based classification and semantic segmentation show that PTQ4SNN effectively preserves model accuracy under W4 quantization and approximately 4-bit membrane precision.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。