Research map / Architectures and efficiency / RNNs and normalization
Implicit Bias of Linear RNNs
Melikasadat Emami, Mojtaba Sahraee-Ardakan, Parthe Pandit, Sundeep Rangan, Alyson K Fletcher
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 |
|---|---|---|---|
| Use the short-memory-biased RNN parameterization instead of unscaled Conv-1D | Macaque S1 reaching models · Limited data (4.5 minutes), x direction | R²: 0.6565 -> 0.6462 | Hurt |
| Use the short-memory-biased RNN parameterization instead of unscaled Conv-1D | Macaque S1 reaching models · Limited data (4.5 minutes), y direction | R²: 0.6027 -> 0.5911 | Hurt |
| Use the short-memory-biased RNN parameterization instead of unscaled Conv-1D | Macaque S1 reaching models · Entire recording, x direction | R²: 0.5856 -> 0.6043 | Improved |
| Use the short-memory-biased RNN parameterization instead of unscaled Conv-1D | Macaque S1 reaching models · Entire recording, y direction | R²: 0.3918 -> 0.4257 | Improved |
| Increase delay in the target system | RNN and scaled Conv-1D · Synthetic delayed-input task | Test error | Hurt |
| Use the short-memory-biased RNN parameterization instead of unscaled Conv-1D | Macaque S1 reaching models · Limited data (4.5 minutes), x and y directions | R²: RNN: 0.6462, 0.5911; Conv-1D: 0.6565, 0.6027 | Mixed |
Closest papers by meaning
- Improved memory in recurrent neural networks with sequential non-normal dynamics (ICLR 2020 · similarity 0.89)
- Recurrent neural networks: vanishing and exploding gradients are not the end of the story (NeurIPS 2024 · similarity 0.89)
- A unified theory of feature learning in RNNs and DNNs (ICML 2026 · similarity 0.88)
- Learning dynamics in linear recurrent neural networks (ICML 2025 · similarity 0.88)
- Training biologically plausible recurrent neural networks on cognitive tasks with long-term dependencies (NeurIPS 2023 · similarity 0.88)
- On the Curse of Memory in Recurrent Neural Networks: Approximation and Optimization Analysis (ICLR 2021 · 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.
