Research map / Probabilistic and scientific modeling / Dynamical systems
Universality and individuality in neural dynamics across large populations of recurrent networks
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 |
|---|---|---|---|
| Change RNN architecture or activation function | Trained RNNs · 3-bit flip-flop | representational geometry similarity | Mixed |
| Compare fixed-point transition structure rather than precise trajectory geometry | Trained RNNs · 3-bit flip-flop | fixed-point topology similarity | Helped |
| Change RNN architecture or activation function | Trained RNNs · Sine wave generation | representational geometry similarity | Hurt |
| Compare fixed-point topology across architectures and activations | Trained RNNs · Sine wave generation | fixed-point topology similarity | No clear effect |
| Use linearized dynamics around input-dependent fixed points | Trained RNNs · Sine wave generation | oscillatory mode frequency alignment | Helped |
| Change RNN architecture or activation function | Trained RNNs · Context-dependent integration | representational geometry similarity | Hurt |
| Compare line-attractor topology rather than representational geometry | Trained RNNs · Context-dependent integration | fixed-point topology similarity | Mixed |
| Use gated architectures instead of a Vanilla RNN | RNNs performing context-dependent integration · Context-dependent integration | number of linear modes used for integration | Mixed |
| Compare trained and untrained networks using SVCCA | Vanilla RNNs · Context-dependent integration | SVCCA similarity | Mixed |
Closest papers by meaning
- Measuring and Controlling Solution Degeneracy across Task-Trained Recurrent Neural Networks (NeurIPS 2025 · similarity 0.89)
- Inference of Neural Dynamics Using Switching Recurrent Neural Networks (NeurIPS 2024 · similarity 0.88)
- Learning dynamics in linear recurrent neural networks (ICML 2025 · similarity 0.88)
- Neuron-based Multifractal Analysis of Neuron Interaction Dynamics in Large Models (ICLR 2025 · similarity 0.88)
- Dynamical phases of short-term memory mechanisms in RNNs (ICML 2025 · similarity 0.88)
- RNNs perform task computations by dynamically warping neural representations (NeurIPS 2025 · 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.