Research map / Architectures and efficiency / Spiking networks
LIF Recurrent Memory Enables Long-Horizon Spiking Computation
Fenghao Liu, Yipeng Shen, Peng Chen, Qian Zheng, Peng Lin, Gang Pan
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add recurrent feedback from context neuron NC to memory neuron NM (ablation) | LRMM · PS-MNIST | accuracy: 86.11 -> 96.52 | Improved |
| Add recurrent feedback from context neuron NC to memory neuron NM (ablation) | LRMM · SHD | accuracy: 86.28 -> 94.70 | Improved |
| Add global output recurrence from NO to the current layer (ablation) | LRMM · PS-MNIST | accuracy: 93.74 -> 96.52 | Improved |
| Use three separate modulatory gates instead of one shared static gate (ablation) | LRMM · PS-MNIST | accuracy: 92.35 -> 96.52 | Improved |
| Use three separate modulatory gates instead of one shared static gate (ablation) | LRMM · SHD | accuracy: 90.81 -> 94.70 | Improved |
| Use all three modulatory currents (II, IF, IO) (ablation) | LRMM · supplementary ablations (Appendix Tables 7-9) | accuracy | Improved |
| Vary the leak factor over a broad range (ablation) | LRMM · supplementary ablations | accuracy | No clear effect |
| Train with modulation-function annealing instead of removing modulation (ablation) | LRMM · supplementary ablations | accuracy and stability | Improved |
| Replace NI and NO with ALIF neurons | LRMM · PS-MNIST | accuracy: 96.52 -> 97.39 | Improved |
| Use LRMM memory loop instead of a much larger ALIF baseline | copy task, delay 20 · copy task L=8, K=8 | accuracy: 0.58 -> 0.89 | Improved |
| Use LRMM instead of LSTM for energy | SHD inference · SHD, 2 layers x 512 neurons | energy (nJ): 21145 -> 76.11 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- TC-LIF — TC-LIF: A Two-Compartment Spiking Neuron Model for Long-Term Sequential Modelling (AAAI 2024) · better on 4 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- TC-LIF: A Two-Compartment Spiking Neuron Model for Long-Term Sequential Modelling (AAAI 2024 · similarity 0.91)
- TS-LIF: A Temporal Segment Spiking Neuron Network for Time Series Forecasting (ICLR 2025 · similarity 0.89)
- P-SPIKESSM: HARNESSING PROBABILISTIC SPIKING STATE SPACE MODELS FOR LONG-RANGE DEPENDENCY TASKS (ICLR 2025 · similarity 0.88)
- SpikingSSMs: Learning Long Sequences with Sparse and Parallel Spiking State Space Models (AAAI 2025 · similarity 0.88)
- CLIF: Complementary Leaky Integrate-and-Fire Neuron for Spiking Neural Networks (ICML 2024 · similarity 0.87)
- Neuronal Self-Adaptation Enhances Capacity and Robustness of Representation in Spiking Neural Networks (NeurIPS 2026 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
