arXiv:2510.15542v2 Announce Type: replace-cross 
Abstract: This paper introduces SpikeFit, a novel training method for Spiking Neural Networks (SNNs) that enables efficient inference on neuromorphic hardware, considering all its stringent requirements: the number of neurons and synapses that can fit on a single device, and lower bit-width representations (e.g., 4-bit, 8-bit). Unlike conventional compressing approaches that address only a subset of these requirements (limited numerical precision and limited number of neurons in the network), SpikeFit treats the allowed weights' discrete values themselves as learnable parameters co-optimized with the model, allowing for optimal Clusterization-Aware Training (CAT) of the model's weights at low precision (2-, 4-, or 8-bit) which results in higher network compression efficiency, as well as limiting the number of unique synaptic connections to a value required by neuromorphic processor. This joint optimization allows SpikeFit to find a discrete weight set aligned with hardware constraints, enabling the most complete deployment across a broader range of neuromorphic processors than existing methods of SNN compression support. Moreover, SpikeFit introduces a new hardware-friendly Fisher Spike Contribution (FSC) pruning method showing the state-of-the-art performance. We demonstrate that for spiking neural networks constrained to only four unique synaptic weight values (M = 4), our SpikeFit method not only outperforms state-of-the-art SNNs compression methods and conventional baselines combining extreme quantization schemes and clustering algorithms, but also meets a wider range of neuromorphic hardware requirements and provides the lowest energy use in experiments.

SpikeFit هي طريقة تدريب مبتكرة لشبكات الأعصاب النابضة (SNN) التي تعمل على تحسين نشرها على الأجهزة العصبية. هذه الابتكار مهم لأنه يتناول القيود المحددة لهذا النوع من الأجهزة، بما في ذلك عدد الخلايا العصبية والوصلات، فضلاً عن الحاجة إلى تمثيلات ذات عرض بت منخفض. من خلال تحسين كفاءة SNN، يمكن أن تعزز SpikeFit بشكل كبير أداء الأنظمة العصبية، مما يجعلها أكثر قابلية للتطبيق في التطبيقات الواقعية في الذكاء الاصطناعي وما بعده.

SpikeFit es un método de entrenamiento innovador para Redes Neuronales Espigadas (SNN) que optimiza su implementación en hardware neuromórfico. Esta innovación es importante porque aborda las restricciones específicas de dicho hardware, incluyendo el número de neuronas y sinapsis, así como la necesidad de representaciones de menor ancho de bits. Al mejorar la eficiencia de las SNN, SpikeFit podría aumentar significativamente el rendimiento de los sistemas neuromórficos, haciéndolos más viables para aplicaciones del mundo real en inteligencia artificial y más.

SpikeFit est une méthode d'entraînement révolutionnaire pour les réseaux de neurones à spikes (SNN) qui optimise leur déploiement sur du matériel neuromorphique. Cette innovation est cruciale car elle répond aux contraintes spécifiques de ce type de matériel, y compris le nombre de neurones et de synapses, ainsi que la nécessité de représentations à faible largeur de bits. En améliorant l'efficacité des SNN, SpikeFit pourrait considérablement renforcer la performance des systèmes neuromorphiques, les rendant plus viables pour des applications concrètes en intelligence artificielle et au-delà.

SpikeFit is a groundbreaking training method for Spiking Neural Networks (SNNs) that optimizes their deployment on neuromorphic hardware. This innovation is crucial as it addresses the specific constraints of such hardware, including the number of neurons and synapses, as well as the need for lower bit-width representations. By improving the efficiency of SNNs, SpikeFit could significantly enhance the performance of neuromorphic systems, making them more viable for real-world applications in artificial intelligence and beyond.

SpikeFit: Towards Optimal Deployment of Spiking Networks on Neuromorphic Hardware

arXiv:2601.08447v1 Announce Type: cross 
Abstract: Spike-timing-dependent plasticity (STDP) provides a biologically-plausible learning mechanism for spiking neural networks (SNNs); however, Hebbian weight updates in architectures with recurrent connections suffer from pathological weight dynamics: unbounded growth, catastrophic forgetting, and loss of representational diversity. We propose a neuromorphic regularization scheme inspired by the synaptic homeostasis hypothesis: periodic offline phases during which external inputs are suppressed, synaptic weights undergo stochastic decay toward a homeostatic baseline, and spontaneous activity enables memory consolidation. We demonstrate that this sleep-wake cycle prevents weight saturation while preserving learned structure. Empirically, we find that low to intermediate sleep durations (10-20\% of training) improve stability on MNIST-like benchmarks in our STDP-SNN model, without any data-specific hyperparameter tuning. In contrast, the same sleep intervention yields no measurable benefit for the surrogate-gradient spiking neural network (SG-SNN). Taken together, these results suggest that periodic, sleep-based renormalization may represent a fundamental mechanism for stabilizing local Hebbian learning in neuromorphic systems, while also indicating that special care is required when integrating such protocols with existing gradient-based optimization methods.

تشير دراسة جديدة إلى اقتراح نظام تنظيم منزلي قائم على النوم لاستقرار البلاستيك المعتمد على توقيت النبضات (STDP) في الشبكات العصبية النابضة المتكررة (SNNs). تهدف هذه الطريقة إلى التخفيف من مشكلات مثل النمو غير المحدود للأوزان والنسيان الكارثي من خلال إدخال مراحل غير نشطة حيث تتراجع الأوزان المشبكية نحو خط الأساس المنزلي، مما يعزز من دمج الذاكرة.

Un nuevo estudio propone un esquema de regularización homeostática basado en el sueño para estabilizar la plasticidad dependiente del tiempo de picos (STDP) en redes neuronales recurrentes de picos (SNNs). Este enfoque busca mitigar problemas como el crecimiento ilimitado de pesos y el olvido catastrófico al introducir fases fuera de línea donde los pesos sinápticos decaen hacia una línea base homeostática, mejorando así la consolidación de la memoria.

Une nouvelle étude propose un schéma de régularisation homéostatique basé sur le sommeil pour stabiliser la plasticité dépendante du timing des pics (STDP) dans les réseaux neuronaux à pics récurrents (SNNs). Cette approche vise à atténuer des problèmes tels que la croissance illimitée des poids et l'oubli catastrophique en introduisant des phases hors ligne où les poids synaptiques décroissent vers une ligne de base homéostatique, améliorant ainsi la consolidation de la mémoire.

A new study proposes a sleep-based homeostatic regularization scheme to stabilize spike-timing-dependent plasticity (STDP) in recurrent spiking neural networks (SNNs). This approach aims to mitigate issues such as unbounded weight growth and catastrophic forgetting by introducing offline phases where synaptic weights decay towards a homeostatic baseline, enhancing memory consolidation.

SpikeFit: Towards Optimal Deployment of Spiking Networks on Neuromorphic Hardware

Was this article worth reading? Share it

LucidQuery AI

Chattermate

Brainactive

Hypertune

Graphite Note

Bytefitz

Ready to build your own newsroom?