# Input-Aware Dynamic Timestep Spiking Neural Networks for Efficient In-Memory Computing

arXiv: https://arxiv.org/abs/2305.17346 · HTML: https://mlsanity.com/paper/2305.17346/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- Builds on 13 of its 25 references that are in the map.
- First posted on arXiv 27 May 2023.

## Builds on (most cited first, 13 of 13)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [On Calibration of Modern Neural Networks](https://mlsanity.com/paper/1706.04599/) (2017)
- [Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding](https://mlsanity.com/paper/1510.00149/) (2015)
- [BranchyNet: Fast Inference via Early Exiting from Deep Neural Networks](https://mlsanity.com/paper/1709.01686/) (2017)
- [Spatio-Temporal Backpropagation for Training High-performance Spiking Neural Networks](https://mlsanity.com/paper/1706.02609/) (2017)
- [Going Deeper With Directly-Trained Larger Spiking Neural Networks](https://mlsanity.com/paper/2011.05280/) (2020)
- [Deep Learning in Spiking Neural Networks](https://mlsanity.com/paper/1804.08150/) (2018)
- [Direct Training for Spiking Neural Networks: Faster, Larger, Better](https://mlsanity.com/paper/1809.05793/) (2018)
- [Adaptive Neural Networks for Efficient Inference](https://mlsanity.com/paper/1702.07811/) (2017)
- [Revisiting Batch Normalization for Training Low-latency Deep Spiking Neural Networks from Scratch](https://mlsanity.com/paper/2010.01729/) (2020)
- [Spatio-Temporal Pruning and Quantization for Low-latency Spiking Neural Networks](https://mlsanity.com/paper/2104.12528/) (2021)
- [SATA: Sparsity-Aware Training Accelerator for Spiking Neural Networks](https://mlsanity.com/paper/2204.05422/) (2022)

## Built on by (most cited first, 11 of 11)

- [TIM: An Efficient Temporal Interaction Module for Spiking Transformer](https://mlsanity.com/paper/2401.11687/) (2024)
- [Towards Low-latency Event-based Visual Recognition with Hybrid Step-wise Distillation Spiking Neural Networks](https://mlsanity.com/paper/2409.12507/) (2024)
- [AT-SNN: Adaptive Tokens for Vision Transformer on Spiking Neural Network](https://mlsanity.com/paper/2408.12293/) (2024)
- [Temporal Flexibility in Spiking Neural Networks: Towards Generalization Across Time Steps and Deployment Friendliness](https://mlsanity.com/paper/2503.17394/) (2025)
- [NeuroNAS: Enhancing Efficiency of Neuromorphic In-Memory Computing for Intelligent Mobile Agents through Hardware-Aware Spiking Neural Architecture Search](https://mlsanity.com/paper/2407.00641/) (2024)
- [When In-memory Computing Meets Spiking Neural Networks -- A Perspective on Device-Circuit-System-and-Algorithm Co-design](https://mlsanity.com/paper/2408.12767/) (2024)
- [DS2TA: Denoising Spiking Transformer with Attenuated Spatiotemporal Attention](https://mlsanity.com/paper/2409.15375/) (2024)
- [A 71.2-μW Speech Recognition Accelerator with Recurrent Spiking Neural Network](https://mlsanity.com/paper/2503.21337/) (2025)
- [Topology Optimization of Random Memristors for Input-Aware Dynamic SNN](https://mlsanity.com/paper/2407.18625/) (2024)
- [Adaptive Time-step Training for Enhancing Spike-Based Neural Radiance Fields](https://mlsanity.com/paper/2507.23033/) (2025)
- [STAS: Spatio-Temporal Adaptive Computation Time for Spiking Transformers](https://mlsanity.com/paper/2508.14138/) (2025)

## Abstract (arXiv)

Spiking Neural Networks (SNNs) have recently attracted widespread research interest as an efficient alternative to traditional Artificial Neural Networks (ANNs) because of their capability to process sparse and binary spike information and avoid expensive multiplication operations. Although the efficiency of SNNs can be realized on the In-Memory Computing (IMC) architecture, we show that the energy cost and latency of SNNs scale linearly with the number of timesteps used on IMC hardware. Therefore, in order to maximize the efficiency of SNNs, we propose input-aware Dynamic Timestep SNN (DT-SNN), a novel algorithmic solution to dynamically determine the number of timesteps during inference on an input-dependent basis. By calculating the entropy of the accumulated output after each timestep, we can compare it to a predefined threshold and decide if the information processed at the current timestep is sufficient for a confident prediction. We deploy DT-SNN on an IMC architecture and show that it incurs negligible computational overhead. We demonstrate that our method only uses 1.46 average timesteps to achieve the accuracy of a 4-timestep static SNN while reducing the energy-delay-product by 80%.