Temporal alignment and latent Gaussian process factor inference in population spike trains

Shared variability is separated into terms that express condition dependence, as well as trial-to-trial variation in trajectories.