Dynamic predictive coding: A model of hierarchical sequence learning and prediction in the neocortex.

We introduce dynamic predictive coding, a hierarchical model of spatiotemporal prediction and sequence learning in the neocortex. The model assumes that higher cortical levels modulate the temporal dynamics of lower levels, correcting their predictions of dynamics using prediction errors. As a result, lower levels form representations that encode sequences at shorter timescales (e.g., a single step) while higher levels form representations that encode sequences at longer timescales (e.g., an entire sequence). We tested this model using a two-level neural network, where the top-down modulation creates low-dimensional combinations of a set of learned temporal dynamics to explain input sequences. When trained on natural videos, the lower-level model neurons developed space-time receptive fields similar to those of simple cells in the primary visual cortex while the higher-level responses spanned longer timescales, mimicking temporal response hierarchies in the cortex. Additionally, the network's hierarchical sequence representation exhibited both predictive and postdictive effects resembling those observed in visual motion processing in humans (e.g., in the flash-lag illusion). When coupled with an associative memory emulating the role of the hippocampus, the model allowed episodic memories to be stored and retrieved, supporting cue-triggered recall of an input sequence similar to activity recall in the visual cortex. When extended to three hierarchical levels, the model learned progressively more abstract temporal representations along the hierarchy. Taken together, our results suggest that cortical processing and learning of sequences can be interpreted as dynamic predictive coding based on a hierarchical spatiotemporal generative model of the visual world.

Expressive probabilistic sampling in recurrent neural networks.

In sampling-based Bayesian models of brain function, neural activities are assumed to be samples from probability distributions that the brain uses for probabilistic computation. However, a comprehensive understanding of how mechanistic models of neural dynamics can sample from arbitrary distributions is still lacking. We use tools from functional analysis and stochastic differential equations to explore the minimum architectural requirements for $\textit{recurrent}$ neural circuits to sample from complex distributions. We first consider the traditional sampling model consisting of a network of neurons whose outputs directly represent the samples (sampler-only network). We argue that synaptic current and firing-rate dynamics in the traditional model have limited capacity to sample from a complex probability distribution. We show that the firing rate dynamics of a recurrent neural circuit with a separate set of output units can sample from an arbitrary probability distribution. We call such circuits reservoir-sampler networks (RSNs). We propose an efficient training procedure based on denoising score matching that finds recurrent and output weights such that the RSN implements Langevin sampling. We empirically demonstrate our model's ability to sample from several complex data distributions using the proposed neural dynamics and discuss its applicability to developing the next generation of sampling-based brain models.

Unsupervised Sleep and Wake State Identification in Long-Term Electrocorticography Recordings.

Studying the neural correlates of sleep can lead to revelations in our understanding of sleep and its interplay with different neurological disorders. Sleep research relies on manual annotation of sleep stages based on rules developed for healthy adults. Automating sleep stage annotation can expedite sleep research and enable us to better understand atypical sleep patterns. Our goal was to create a fully unsupervised approach to label sleep and wake states in human electro-corticography (ECoG) data from epilepsy patients. Here, we demonstrate that with continuous data from a single ECoG electrode, hidden semi-Markov models (HSMM) perform best in classifying sleep/wake states without excessive transitions, with a mean accuracy (n=4) of 85.2% compared to using K-means clustering (72.2%) and hidden Markov models (81.5%). Our results confirm that HSMMs produce meaningful labels for ECoG data and establish the groundwork to apply this model to cluster sleep stages and potentially other behavioral states.