Latent Dynamics Models of Neural Population Recordings
Sequence models that infer low-dimensional dynamics from many simultaneously recorded neurons.
Tier Grades
Tier A: Frontier AI labs
feasible
Google with Stanford published LFADS.
Tier B: Indonesian AI labs
stretch
The model is small, but no Indonesian neuroscience recording programme is observed.
Tier C: Institution units such as ITB
stretch
The model is small, but it needs multi-electrode neural recordings.
Tier D: Organisations such as SEMAI (PoT, PoC, PoV)
not feasible
Neural recording analysis has no delivery use case in the list.
Where this comes from
Tiers: a) frontier AI labs, b) Indonesian AI labs, c) institution units such as ITB, d) organisations such as SEMAI delivering proof of technology, concept and value. Each rationale cites the anchors below.
Anchors
- ComputeLFADS: a variational auto-encoder; compute not stated in the abstract.
- DataLFADS: single-trial, high-dimensional neural spiking data and synthetic data.
- Team size (proxy)Proxy only (author count of an exemplar paper, not a headcount): 4 authors on the LFADS paper (arXiv 1608.06315).
- Time horizonYears.
- Shelf-life signalsnot in the notes
- Pipeline roleTier A and universities publish methods; Tier C applies them where recordings exist; Tier B and Tier D have no observed role.
Where this comes from
Team size is a proxy (author counts of exemplar papers), not a headcount.
Who and What
Links
R&D TopicLatent Dynamics Models of Neural Population Recordings
Models
LFADSGoogle, scientific
LFADSGoogle, scientificUse Casesnone recorded
Who Runs Itnone recorded
Papers
- A Unified, Scalable Framework for Neural Population DecodingMehdi Azabou et al., NeurIPS 2023 (per arXiv comment), 2023 state of the art
- Neuroformer: Multimodal and Multitask Generative Pretraining for Brain DataAntonis Antoniades et al., arXiv preprint, 2023 state of the art
- Representation learning for neural population activity with Neural Data TransformersJoel Ye et al., arXiv preprint, 2021 state of the art
- Neural Latents Benchmark '21: Evaluating latent variable models of neural population activityFelix Pei et al., arXiv preprint, 2021 state of the art
Where this comes from
The papers to read first: where the direction began, its current best, and a survey when one exists.
Sources
- arxiv.org/abs/1608.06315 https://arxiv.org/abs/1608.06315
To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Latent Dynamics Models of Neural Population Recordings", https://aurion.semai.id/topics/latent-dynamics-of-neural-recordings/, records of October 2026.