Research map / Probabilistic and scientific modeling / Dynamical systems
Reverse engineering recurrent neural networks with Jacobian switching linear dynamical systems
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 |
|---|---|---|---|
| Co-train with the JSLDS and expansion network | GRU on 3-bit memory · 3-bit memory | mean relative error | Helped |
| Co-train with the JSLDS and expansion network | Vanilla RNN · Context-dependent integration | mean relative error | Helped |
| Use JSLDS co-training | GRU on 3-bit memory · 3-bit memory | fixed-point structure | Mixed |
| Use JSLDS co-training (ablation) | Vanilla RNN · Context-dependent integration | fixed-point solution | No clear effect |
| Use JSLDS co-training (ablation) | LFADS generator · Monkey reach task | fixed-point solution | No clear effect |
| Set the RNN loss weight to 3.0 while keeping the JSLDS loss weight at 1.0 (ablation) | JSLDS-RNN · 3-bit memory | task performance | Helped |
Closest papers by meaning
- Inference of Neural Dynamics Using Switching Recurrent Neural Networks (NeurIPS 2024 · similarity 0.92)
- Learning Dynamics of RNNs in Closed-Loop Environments (NeurIPS 2025 · similarity 0.90)
- Modeling Latent Neural Dynamics with Gaussian Process Switching Linear Dynamical Systems (NeurIPS 2024 · similarity 0.88)
- Mechanistic Interpretability of RNNs emulating Hidden Markov Models (NeurIPS 2025 · similarity 0.88)
- Learning dynamics in linear recurrent neural networks (ICML 2025 · similarity 0.88)
- Almost-Linear RNNs Yield Highly Interpretable Symbolic Codes in Dynamical Systems Reconstruction (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.