Research map / Architectures and efficiency / Spiking networks
Learning to Time-Decode in Spiking Neural Networks Through the Information Bottleneck
Nicolas Skatchkovsky, Osvaldo Simeone, Hyeryung Jang
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 |
|---|---|---|---|
| VDIB learned time decoding instead of rate decoding | VDIB, MLP decoder · MNIST, rate encoding | accuracy: 60.50 ± 0.32 % -> 91.32 ± 0.27 % | Improved |
| VDIB learned time decoding instead of rate decoding | VDIB, MLP decoder · MNIST, time-to-first-spike encoding | accuracy: 86.43 ± 0.84 % -> 92.82 ± 0.25 % | Improved |
| VDIB instead of rate decoding | VDIB, conv decoder · MNIST-DVS | reconstruction MSE: 0.19 ± 0.02 -> 0.02 ± 0.0006 | Improved |
| Convolutional decoder instead of MLP decoder (ablation) | VDIB · MNIST, rate encoding | accuracy: 91.32 ± 0.27 % -> 88.73 ± 1.06 % | Hurt |
| Convolutional decoder instead of MLP decoder (ablation) | VDIB · MNIST, time-to-first-spike encoding | accuracy: 92.82 ± 0.25 % -> 90.72 ± 0.39 % | Hurt |
| Time-to-first-spike encoding instead of rate encoding of inputs | VDIB, MLP decoder · MNIST | accuracy: 91.32 ± 0.27 % -> 92.82 ± 0.25 % | Improved |
| Larger encoding and decoding windows tau_e and tau_d (ablation) | VDIB · Predictive coding (filtering) | MSE | Improved |
| Encoding window tau_e > 3 (ablation) | VDIB · Predictive coding (filtering) | MSE | Improved |
| Increase regularization strength beta (ablation) | VDIB · Predictive coding | spike rate vs accuracy | Mixed |
Closest papers by meaning
- Efficiently Training Time-to-First-Spike Spiking Neural Networks from Scratch (ICML 2026 · similarity 0.90)
- IM-Loss: Information Maximization Loss for Spiking Neural Networks (NeurIPS 2022 · similarity 0.90)
- EnOF-SNN: Training Accurate Spiking Neural Networks via Enhancing the Output Feature (NeurIPS 2024 · similarity 0.89)
- A Scalable, Causal, and Energy Efficient Framework for Neural Decoding with Spiking Neural Networks (NeurIPS 2025 · similarity 0.89)
- Advancing Training Efficiency of Deep Spiking Neural Networks through Rate-based Backpropagation (NeurIPS 2024 · similarity 0.88)
- Take A Shortcut Back: Mitigating the Gradient Vanishing for Training Spiking Neural Networks (NeurIPS 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.
