Research map / Probabilistic and scientific modeling / Dynamical systems
Low-rank and diffusion neural dynamics
Extracts neural population dynamics using low-rank RNNs, latent diffusion and nonlinear embeddings.
53 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · an approach in Dynamical systems, keyword cluster “brain · activity · neurons”
State History Tried Checklist Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
11 papers in 2023–24 and 18 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic held steady, from 4.9% to 4.7%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2016 · Bayesian latent structure discovery from multi-neuron recordings (NeurIPS 2016 · 44 citations)
- 2017 · Gaussian process based nonlinear latent structure discovery in multivariate spike train data (NeurIPS 2017 · 81 citations)
- 2018 · Temporal alignment and latent Gaussian process factor inference in population spike trains (NeurIPS 2018 · 42 citations)
- 2019 · Enabling hyperparameter optimization in sequential autoencoders for spiking neural data (NeurIPS 2019 · 22 citations)
- 2020 · Identifying signal and noise structure in neural population activity with Gaussian process factor models (NeurIPS 2020 · 16 citations)
- 2021 · Inferring Latent Dynamics Underlying Neural Population Activity via Neural Differential Equations (ICML 2021 · 20 citations)
- 2022 · Extracting computational mechanisms from neural data using low-rank RNNs (NeurIPS 2022 · 12 citations)
- 2023 · Understanding Neural Coding on Latent Manifolds by Sharing Features and Dividing Ensembles (ICLR 2023 · 5 citations)
- 2024 · Inferring stochastic low-rank recurrent neural networks from neural data (NeurIPS 2024 · 4 citations)
- 2025 · Flow-field inference from neural data using deep recurrent networks (ICML 2025 · 9 citations)
- 2026 · Coupled Transformer Autoencoder for Disentangling Multi-Region Neural Latent Dynamics (ICLR 2026 · 1 citation)
Earliest papers here
- Bayesian latent structure discovery from multi-neuron recordings (NeurIPS 2016 · 44 citations)
- Linear dynamical neural population models through nonlinear embeddings (NeurIPS 2016 · 41 citations)
- Learning Population-Level Diffusions with Generative RNNs (ICML 2016 · 34 citations)
What has been tried
38 papers here report 322 measured changes (a component added, removed or swapped, compared with a baseline): helped 61%, no clear effect 15%, hurt 21%, mixed 4%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| Add a task-irrelevant time-varying component | Rat frontal cortex decision-making dataset · dynamical representation consistency and interpretability | Helped | Flow-field inference from neural data using deep recurrent networks (ICML 2025 · 9 citations) | |
| use a non-causal CNN encoder instead of a causal encoder | MC_Maze · co-bps | 0.3225 -> 0.3407 | Helped | Inferring stochastic low-rank recurrent neural networks from neural data (NeurIPS 2024 · 4 citations) |
| add a spike-history autoregressive observation model | Monkey reach · D_KL population spike count histogram | 0.0039 ± 3.9\text{e-}4 -> 0.0016 ± 6.2\text{e-}4 | Helped | Latent Diffusion for Neural Spiking Data (NeurIPS 2024 · 3 citations) |
| prune high Fourier frequencies | Simulated shared and independent latents · latent identification accuracy | Helped | Multi-modal Gaussian Process Variational Autoencoders for Neural and Behavioral Data (ICLR 2024 · 2 citations) | |
| remove bottom 95% of recurrent weights | Quenched RNN · critical gain | No clear effect | Slow Transition to Low-Dimensional Chaos in Heavy-Tailed Recurrent Neural Networks (NeurIPS 2025 · 1 citation) | |
| Use multiscale encoder rather than zero-imputation for missing samples | NHP reaching datasets · behavior decoding accuracy | Helped | Dynamical modeling of nonlinear latent factors in multiscale neural activity with real-time inference (NeurIPS 2025 · 0 citations) | |
| Vary the generalized Laguerre basis hyperparameters c and α | Simulated data · filter MSE | No clear effect | Scalable inference of functional neural connectivity at submillisecond timescales (NeurIPS 2025 · 0 citations) | |
| Infer Poisson spike counts by rounding the model output rather than sampling | Chicken RGC test set · spike-train similarity | Helped | Spike Distance Function as a Learning Objective for Spike Prediction (ICML 2024 · 0 citations) |
What reviewers will expect
Datasets and baselines named in the 15 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Baselines
- CEBRA · 2 of 15 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 29 papers; from OpenAlex and the papers' affiliations).
- New York University · 3 papers
- University of Cambridge · 2 papers
- Fordham University · 2 papers
- Georgia Institute of Technology · 2 papers
- Allen Institute · 2 papers
- University of Washington · 2 papers
- Princeton University · 2 papers
- Eberhard-Karls-Universität Tübingen · 2 papers
- University of Tuebingen · 2 papers
- Flatiron Institute · 2 papers
Use it
Download BibTeX of 39 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Latent models of neural activity (70 papers)
- Recurrent circuit models of inference (70 papers)
- Functional connectivity modeling (50 papers)
- Synaptic plasticity learning rules (72 papers)
Elsewhere in the topic
- Physics-informed dynamics learning (127 papers)
- Latent dynamics models (111 papers)
- State-space inference for stochastic systems (104 papers)
- Switching and law-discovery dynamics (81 papers)
- Latent-space control and planning (78 papers)
- Data assimilation from sparse observations (58 papers)
- Stochastic population dynamics (53 papers)
- Stable contractive dynamics (42 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).