Research map / Architectures and efficiency / Spiking networks
SEENN: Towards Temporal Spiking Early Exit Neural Networks
Yuhang Li, Tamar Geller, Youngeun Kim, Priyadarshini Panda
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 |
|---|---|---|---|
| replace fixed timesteps with confidence-threshold early exit (SEENN-I) | directly trained ResNet-19 · CIFAR-10 | accuracy at avg timesteps: 95.45 (T=2) -> 96.07 (1.09) | Improved |
| SEENN-I applied to ANN-SNN conversion instead of fixed T | QCFS ResNet-18 · CIFAR-10 | accuracy: 75.44 (T=2) -> 93.63 (1.4) | Improved |
| use RL policy (SEENN-II) instead of confidence thresholding (SEENN-I) (ablation) | SEENN · CIFAR-100 ResNet-19 | accuracy: 79.56 -> 80.23 | Improved |
| use RL policy (SEENN-II) instead of confidence thresholding (ablation) | SEENN · CIFAR-10 ResNet-19 | accuracy: 96.07 -> 96.01 | No clear effect |
| SEENN-II instead of SEENN-I on ImageNet SEW-ResNet-34 (ablation) | SEENN · ImageNet | avg timesteps: 2.35 -> 1.79 | Improved |
| early exit on directly trained SNN | TET ResNet-34 ImageNet · ImageNet | accuracy: 64.79 (6) -> 64.66 (3.38) | No clear effect |
| early exit on CIFAR10-DVS | VGG-SNN · CIFAR10-DVS | accuracy: 83.1 -> 82.7 | Hurt |
| raise confidence threshold alpha (ablation) | SEENN-I · ImageNet SEW-ResNet-34 | share of inputs exiting at 1 timestep: 73.4% -> 30.7% | Mixed |
| add per-timestep cross-entropy (TET loss) in SEENN-I training (ablation) | SEENN-I · CIFAR/ImageNet | accuracy | Improved |
Closest papers by meaning
- TS-SNN: Temporal Shift Module for Spiking Neural Networks (ICML 2025 · similarity 0.89)
- Temporal Misalignment in ANN-SNN Conversion and its Mitigation via Probabilistic Spiking Neurons (ICML 2025 · similarity 0.89)
- Temporal Flexibility in Spiking Neural Networks: Towards Generalization Across Time Steps and Deployment Friendliness (ICLR 2025 · similarity 0.89)
- EnOF-SNN: Training Accurate Spiking Neural Networks via Enhancing the Output Feature (NeurIPS 2024 · similarity 0.89)
- Efficient Logit-based Knowledge Distillation of Deep Spiking Neural Networks for Full-Range Timestep Deployment (ICML 2025 · similarity 0.89)
- Sign Gradient Descent-based Neuronal Dynamics: ANN-to-SNN Conversion Beyond ReLU Network (ICML 2024 · similarity 0.88)
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.
