The paper investigates the expressivity of time-to-first-spike spiking neural networks, showing that each neuron's firing time can be represented in a maxout-like form with many constrained affine pieces. It formalizes causal regions as polyhedral sets defined by fixed causal spike sequences and derives bounds on the number of such regions for both shallow and multilayer networks. Experiments confirm that spiking networks can produce richer input-space partitions than conventional feedforward ReLU networks.
arXiv:2504.14015v2 Announce Type: replace-cross Abstract: We introduce "causal pieces", a novel concept for analysing spiking neural networks (SNNs), inspired by "linear pieces" used to study express...
arXiv:2605. 08022v2 Announce Type: replace-cross Abstract: Spiking Neural Networks (SNNs) have been proposed as biologically plausible and energy-efficient alternatives to conventional Artificial Neural Networks (ANNs).
arXiv:2607. 14672v1 Announce Type: new Abstract: Continuous-time spiking neural networks (SNNs) provide an event-driven framework for temporal computation, computational neuroscience, and neuromorphic hardware.
The paper demonstrates that for a broad class of spiking neuron models, including the leaky integrate‑and‑fire with subtractive reset, any approximation bound proven for multi‑spike networks can be translated to an equivalent single‑spike network with only a linear change in neuron count, and vice versa. This establishes that single‑spike and multi‑spike neural networks possess identical approximation capabilities for general machine learning tasks. Consequently, existing approximation results for single‑spike networks automatically extend to the multi‑spike case.
The paper introduces SpiKFAX, a second‑order optimization technique for Spiking Neural Networks (SNNs) that uses a Kronecker‑factored approximation of the Fisher information matrix tailored to the sparse, discrete, and temporally recurrent dynamics of SNNs. By addressing the sharp loss landscape that hampers training with conventional optimizers, SpiKFAX improves test accuracy and training stability across five architectures and seven datasets. The method offers a computationally tractable alternative to existing curvature‑based approaches for SNNs.
arXiv:2512. 01906v3 Announce Type: replace Abstract: Spiking neural networks (SNNs) are biologically inspired, event-driven models suited for temporal data processing and energy-efficient neuromorphic computing.
arXiv:2608.29869v1 Announce Type: new Abstract: Spiking neural networks (SNNs) are often regarded as energy-efficient alternatives to artificial neural networks (ANNs), yet their advantage depends cr...
arXiv:2610. 22484v1 Announce Type: cross Abstract: The input space of single-spike neural networks decomposes into causal pieces, and the number of causal pieces has been shown to provide a lower bound on the approximation error for some spiking neural networks.
The paper presents a convex formulation for training spiking neural networks (SNNs), showing that the weight‑decay‑regularized training problem for a class of threshold RNNs—including leaky integrate‑and‑fire SNNs—has zero duality gap and matches a convex program’s optimal value. To handle the exponential size of the resulting convex programs, the authors introduce a randomized surrogate with a high‑probability approximation guarantee. Empirical results demonstrate that the convex variant outperforms surrogate‑gradient training on addition and first‑last‑XOR tasks, matches performance on sequential MNIST, and falls slightly behind a reservoir baseline on base‑3 and base‑5 addition.
The paper introduces the Dormant Spiking Neuron (DSN), a state-driven mechanism that allows neurons in spiking neural networks to enter a dormant state when their membrane potential falls below a negative threshold. This dynamic gating prunes redundant computations at the instance level and is trained end‑to‑end via backpropagation through time with surrogate gradients. Experiments on static and neuromorphic datasets, including ImageNet, show that DSN reduces computational energy by 40% to over 70% and improves baseline SNN accuracy by 1.34%, outperforming state‑of‑the‑art static pruning methods.
arXiv:2606. 03935v1 Announce Type: cross Abstract: The ability to train spiking neural networks is essential for modeling biological neural networks as well as for neuromorphic computing.
arXiv:2610. 33658v1 Announce Type: new Abstract: Vision-language-action (VLA) models have demonstrated strong capabilities in robotic manipulation, yet their reliance on dense, continuous-valued computation presents a fundamental challenge for energy-constrained deployment.