mlsanityData to 7 Oct 2026

SPARQ: Spiking Early-Exit Neural Networks for Energy-Efficient Edge AI

arXiv 2603.14380 ↗First posted 15 Mar 2026cs.LG cs.AI cs.ARcited by 1

On the map: Efficient LLM inference and serving → Early exit and edge inference

Abstract (arXiv)

Spiking neural networks (SNNs) offer inherent energy efficiency due to their event-driven computation model, making them promising for edge AI deployment. However, their practical adoption is limited by the computational overhead of deep architectures and the absence of input-adaptive control. This work presents SPARQ, a unified framework that integrates spiking computation, quantization-aware training, and reinforcement learning-guided early exits for efficient and adaptive inference. Evaluations across MLP, LeNet, and AlexNet architectures demonstrated that the proposed Quantised Dynamic SNNs (QDSNN) consistently outperform conventional SNNs and QSNNs, achieving up to 5.15% higher accuracy over QSNNs, over 330 times lower system energy compared to baseline SNNs, and over 90 percent fewer synaptic operations across different datasets. These results validate SPARQ as a hardware-friendly, energy-efficient solution for real-time AI at the edge.

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Placed in Early exit and edge inference, one of 12 directions in Efficient LLM inference and serving. Builds on 14 of its 24 references that are in the map. First posted on arXiv 15 Mar 2026.

Builds on

14 of 24 references.

References · 14

  1. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,155
  2. 2017Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inferencecited by 1,548
  3. 2017BranchyNet: Fast Inference via Early Exiting from Deep Neural Networkscited by 591
  4. 2021Dynamic Neural Networks: A Surveycited by 405
  5. 2018Deep Learning in Spiking Neural Networkscited by 258
  6. 2021Training Spiking Neural Networks Using Lessons From Deep Learningcited by 251
  7. 2020Knowledge Distillation Meets Self-Supervisioncited by 142
  8. 2024Spike-driven Transformer V2: Meta Spiking Neural Network Architecture Inspiring the Design of Next-generation Neuromorphic Chipscited by 97
  9. 2023SEENN: Towards Temporal Spiking Early-Exit Neural Networkscited by 39
  10. 2023Workload-Balanced Pruning for Sparse Spiking Neural Networkscited by 13
  11. 2024Q-SNNs: Quantized Spiking Neural Networkscited by 10
  12. 2023SpikingNeRF: Making Bio-inspired Neural Networks See through the Real Worldcited by 6
2 more
  1. 2024Toward Large-scale Spiking Neural Networks: A Comprehensive Survey and Future Directionscited by 6
  2. 2023Highly Efficient SNNs for High-speed Object Detectioncited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026An FPGA-Based SoC Architecture with a RISC-V Controller for Energy-Efficient Temporal-Coding Spiking Neural Networkscited by 1

Similar, not cited

Close in wording, no citation either way

  1. 2025The Enduring Dominance of Deep Neural Networks: A Critical Analysis of the Fundamental Limitations of Quantum Machine Learning and Spiking Neural Networks
  2. 2025Enabling Efficient Processing of Spiking Neural Networks with On-Chip Learning on Commodity Neuromorphic Processors for Edge AI Systems
  3. 2025SpikingMamba: Towards Energy-Efficient Large Language Models via Knowledge Distillation from Mamba
  4. 2026The Sparsity Ceiling: Where Spiking Networks Can and Cannot Trade Activity for Energy
  5. 2025Spiking Decision Transformers: Local Plasticity, Phase-Coding, and Dendritic Routing for Low-Power Sequence Control
  6. 2025Adaptive Surrogate Gradients for Sequential Reinforcement Learning in Spiking Neural Networks
  7. 2026SpikingVLA: Asynchronous Spiking Vision-Language-Action Models
  8. 2023TopSpark: A Timestep Optimization Methodology for Energy-Efficient Spiking Neural Networks on Autonomous Mobile Agents