This work analyzes flaws of conventional conversion pipelines from residual membrane potential statistics and proposes a novel conversion strategy combining dynamic initial potential tuning and feature enhancement, which generalizes to ReLU CNNs, ANN Transformers, and multi-threshold SNN variants.
Abstract
Spiking Neural Networks (SNNs) serve as core architectures for neuromorphic computing thanks to event-driven operation and ultra-low power consumption. Direct SNN training is hindered by non-differentiable spikes that induce vanishing gradients and unstable optimization. ANN-SNN conversion circumvents such issues by reusing well-trained ANN weights for low-latency, energy-efficient inference. Nevertheless, existing conversion schemes suffer from severe accuracy drops at small timesteps, large inference delays and cumulative quantization errors, even with marginal performance loss at large $T$. To address these limitations, we first analyze flaws of conventional conversion pipelines from residual membrane potential statistics and propose a novel conversion strategy combining dynamic initial potential tuning and feature enhancement. We then introduce a regularization loss $\mathcal{L}_{\mathrm{RMPD}}$ to adapt initial potential of IF neurons and mitigate systematic truncation bias from boundary aggregation. A dedicated SCR-Conv2d competitive refinement layer with grouped convolution is further built to sharpen feature discrimination, eliminate redundant spikes and stabilize encoding under tiny time windows. Integrated with the state-of-the-art QCFS baseline, our approach delivers consistent low-latency performance gains and generalizes to ReLU CNNs, ANN Transformers, and multi-threshold SNN variants. Evaluations on CIFAR-10, CIFAR-100 and ImageNet verify prominent accuracy improvements at $T=2,4,8$, with negligible extra computation overhead. This work offers an effective conversion paradigm to facilitate real-world SNN deployment on neuromorphic chips.
Extensive experiments on static and neuromorphic benchmarks show that lower-bit BASC models match or outperform higher-bit baselines and retain this accuracy advantage after structured pruning, while further reducing model storage and synaptic operations.
Linliang Chen, Yan Zhong, Xin Liu et al.· 0 citations
PTQ4SNN is proposed, a membrane-aware post-training quantization framework that jointly quantizes weights and recurrent membrane states using only a small calibration set and effectively preserves model accuracy under W4 quantization and approximately 4-bit membrane precision.
Hui Xie, Tong Shi, Haotong Qin et al.· 0 citations
Spiking neural networks (SNNs) offer an energy-efficient alternative to conventional deep neural networks by exploiting sparse event-driven computation, but their training remains challenging because the non-differentiable spike function requires surrogate gradients whose fixed shape may be suboptimal across layers and training stages. In this work, we introduce SAGE, an uncertainty-modulated surrogate-gradient mechanism for Transformer-based SNNs. SAGE estimates block-level uncertainty from normalized self-attention entropy and uses this signal to adapt the surrogate-gradient slope during training while leaving the inference model unchanged. By modulating only the training-time surrogate parameter, the proposed method preserves the original architecture and deployment cost while improving optimization flexibility. Experiments on CIFAR-10/100 demonstrate that SAGE achieves improved accuracy over fixed-surrogate baselines, with results up to 1-2\% consistent gains across multiple simulation time steps. These results highlight the potential of attention-derived uncertainty as a lightweight training signal for adaptive surrogate-gradient learning in transformer-based SNNs.
Spiking Neural Networks (SNNs) have emerged as an energy-efficient alternative to Artificial Neural Networks (ANNs), leveraging sparse accumulate operations in the place of power-hungry multiply-and-accumulate operations. ANN-SNN conversion is a widely adopted approach to realize deep SNNs with accuracy comparable to that of ANNs. The Quantization-Clip-Floor-Shift (QCFS) activation minimizes conversion error, yet requires a large number of inference timesteps to match the source ANN accuracy on real-world vision datasets. PASCAL addresses this by proposing the Precise ANN-SNN Conversion Integrate-and-Fire (PASC-IF) neuron, which guarantees mathematical equivalence between the converted SNN and the source ANN, thereby achieving ANN-equivalent accuracy at significantly reduced timesteps. Despite this algorithmic advancement, the hardware implications of deploying the PASC-IF neuron remain unexplored. In this work, we present APEX, a dual-sparsity SNN inference accelerator that integrates the PASC-IF neuron into the LoAS hardware framework. The three-stage PASC-IF datapath is realized as a fully combinational circuit with no additional latency cost. APEX exploits dual sparsity in both input spikes and weights through a fully temporal-parallel dataflow, enabling efficient sparse computation and reduced memory traffic. Across all evaluated models, the PASC-IF neuron on average achieves up to 3% higher accuracy than the standard IF neuron, with a power overhead of only 1.3%-5.4%, an area overhead of 2.1%-2.7%, and 40% energy reduction for best accuracy configurations.
Adaptive Fission is proposed, a post-training encoding technique that selectively splits high-sensitivity neurons into groups with varying scales and weights that enables neuron-specific, on-demand precision and threshold allocation while introducing minimal spatial overhead.
Yizhou Jiang, Feng Chen, Yihan Li et al.· Neural Information Processin...· 2 citations
The proliferation of artificial neural networks (ANNs) has led to increased energy consumption, raising concerns about their sustainability. Spiking neural networks (SNNs) operate using sparse, binary spikes to communicate information between neurons and offer a potential solution due to their limited energy requirements. Another technique for reducing a neural network's footprint is quantization, which compresses weight representations to decrease memory usage and energy consumption. In this study, we present twin network augmentation (TNA), a novel training method aimed at improving the performance of SNNs on a range of benchmark data sets while also facilitating enhanced network compression through quantization of weights. TNA involves cotraining an SNN with a twin SNN with identical network architecture, optimizing both networks to minimize their cross-entropy losses and the mean squared error between their output logits. We demonstrate that TNA significantly enhances classification performance across various data sets and can be applied for reducing SNNs to ternary weight precision for inference. Our results show that TNA outperforms traditional knowledge distillation methods and achieves state-of-the-art performance for the evaluated network architecture on benchmark data sets, including CIFAR-10, CIFAR-100, and CIFAR-10-DVS. This letter underscores the effectiveness of TNA in bridging the performance gap between SNNs and ANNs and suggests further exploration into the application of TNA on different network architectures and data sets.
Lucas Deckers, Benjamin Vandersmissen, Ing Jyh Tsang et al.· Neural Computation· 0 citations
