Research map / Probabilistic and scientific modeling / Dynamical systems
Extracting low-dimensional dynamics from multiple large-scale neural population recordings by learning to predict correlations
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 |
|---|---|---|---|
| Use dynamics-agnostic S3ID rather than FA with post-hoc alignment | Zebrafish imaging data · Held-out voxel pairs | instantaneous covariance correlation: 0.71 | Helped |
| Use dynamics-agnostic S3ID rather than GROUSE | Zebrafish imaging data · Held-out voxel pairs | instantaneous covariance correlation: 0.72 | Helped |
| Use nonlinear rather than linear latent dynamics (ablation) | S3ID on the spiking neural network · Non-observed pairwise covariances | time-lagged covariance prediction | Helped |
| Use nonlinear rather than linear latent dynamics (ablation) | S3ID on the zebrafish imaging data · Held-out voxel pairs | covariance reconstruction | Helped |
| Initialize sEM with S3ID rather than random parameters | Stitching LDS models · Simulated LDS at multiple population sizes | subspace estimation performance | Helped |
| Set latent dimensionality to 5 rather than the true value 10 (ablation) | S3ID · Simulated LDS, 5% overlap | unobserved instantaneous covariance correlation: 0.99 -> 0.69 | Hurt |
| Set latent dimensionality to 8 rather than the true value 10 (ablation) | S3ID · Simulated LDS, 5% overlap | unobserved instantaneous covariance correlation: 0.99 -> 0.89 | Hurt |
| Set latent dimensionality to 20 rather than the true value 10 (ablation) | S3ID · Simulated LDS, 5% overlap | unobserved instantaneous covariance correlation: 0.99 -> 0.97 | Hurt |
| Set latent dimensionality to 50 rather than the true value 10 (ablation) | S3ID · Simulated LDS, 5% overlap | unobserved instantaneous covariance correlation: 0.99 -> 0.96 | Hurt |
| Reduce cluster neuron sampling from 50% to 40% (ablation) | Dynamics-agnostic S3ID · Spiking neural network, undersampled clusters | instantaneous covariance correlation: 0.97 -> 0.80 | Hurt |
| Reduce cluster neuron sampling from 40% to 30% (ablation) | Dynamics-agnostic S3ID · Spiking neural network, undersampled clusters | instantaneous covariance correlation: 0.80 -> 0.32 | Hurt |
Closest papers by meaning
- CREIMBO: Cross-Regional Ensemble Interactions in Multi-view Brain Observations (ICLR 2025 · similarity 0.87)
- Meta-Dynamical State Space Models for Integrative Neural Data Analysis (ICLR 2025 · similarity 0.87)
- Modeling Latent Neural Dynamics with Gaussian Process Switching Linear Dynamical Systems (NeurIPS 2024 · similarity 0.86)
- Learning Interpretable Hierarchical Dynamical Systems Models from Time Series Data (ICLR 2025 · similarity 0.86)
- Spectral Learning of Shared Dynamics Between Generalized-Linear Processes (NeurIPS 2024 · similarity 0.86)
- Partial observation can induce mechanistic mismatches in data-constrained models of neural dynamics (NeurIPS 2024 · similarity 0.86)
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.