Research map / Architectures and efficiency / RNNs and normalization
Reservoir Computing meets Recurrent Kernels and Structured Transforms
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 dense random reservoir weights with structured Hadamard and random diagonal matrices | Reservoir Computing · Kuramoto-Sivashinsky | forward time: 10.4 -> 3.5 | Helped |
| Replace dense random reservoir weights with structured Hadamard and random diagonal matrices | Reservoir Computing · Kuramoto-Sivashinsky | prediction time: 15.4 -> 3.7 | Helped |
| Replace dense random reservoir weights with structured Hadamard and random diagonal matrices | Reservoir Computing · Kuramoto-Sivashinsky | prediction performance | No clear effect |
| Increase the reservoir size (ablation) | Reservoir Computing and Structured Reservoir Computing · Kuramoto-Sivashinsky | MSE | Helped |
| Use Structured Reservoir Computing instead of conventional Reservoir Computing (ablation) | Reservoir Computing · Recurrent-kernel convergence study | MSE to recurrent-kernel limit | Helped |
| Keep the same random weight matrices across timesteps instead of resampling them (ablation) | Reservoir Computing and Structured Reservoir Computing · Recurrent-kernel convergence study | MSE to recurrent-kernel limit | No clear effect |
| Use recursive prediction by feeding each prediction back into the model (ablation) | Reservoir Computing · Kuramoto-Sivashinsky | future prediction performance | Helped |
| Use direct multi-step prediction instead of recursive prediction (ablation) | Reservoir Computing · Kuramoto-Sivashinsky | future prediction performance | Hurt |
| Use recurrent kernels instead of Reservoir Computing | Reservoir Computing · Kuramoto-Sivashinsky | prediction time: 15.4 -> 23.0 | Hurt |
Closest papers by meaning
- Universality of Real Minimal Complexity Reservoir (AAAI 2025 · similarity 0.90)
- Frequency Domain Reservoir Computing (NeurIPS 2026 · similarity 0.87)
- Spectral-Aware Reservoir Computing for Fast and Accurate Time Series Classification (ICML 2025 · similarity 0.86)
- Accelerating Linear Recurrent Neural Networks for the Edge with Unstructured Sparsity (ICML 2025 · similarity 0.84)
- Recurrent neural networks: vanishing and exploding gradients are not the end of the story (NeurIPS 2024 · similarity 0.84)
- Parallelizing non-linear sequential models over the sequence length (ICLR 2024 · similarity 0.83)
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.