Bubblewrap: Online tiling and real-time flow prediction on neural manifolds
Published in Advances in Neural Information Processing Systems 34, 2021
We present a method that combines fast, stable dimensionality reduction with a soft tiling of the resulting neural manifold, allowing dynamics to be approximated as a probability flow between tiles. This method can be fit efficiently using online expectation maximization, scales to tens of thousands of tiles, and outperforms existing methods when dynamics are noise-dominated or feature multi-modal transition probabilities. The resulting model can be trained at kiloHertz data rates, produces accurate approximations of neural dynamics within minutes, and generates predictions on submillisecond time scales. It retains predictive performance throughout many time steps into the future and is fast enough to serve as a component of closed-loop causal experiments.
Draelos, A., Gupta, P., Jun, N. Y., Sriworarat, C., and Pearson, J. (2021). "Bubblewrap: Online tiling and real-time flow prediction on neural manifolds." Advances in Neural Information Processing Systems 34 (accepted; also at arXiv:2108.13941)