Although behaviorally relevant patterns of neocortical activity in specific frequency bands have been broadly characterized, identifying individual underlying network events remains a key challenge in understanding information processing in cortical circuits. Using a novel analytical method for temporally precise detection of discrete network events, we identified and tracked discrete sets of events underlying two major forms of state-dependent activity patterns in mouse V1 cortex in the β (15-30Hz) and γ (30-80Hz) ranges. γ events regulated spike timing and selectively enhanced visual encoding. Precise tracking revealed that γ, but not β, event rates increased prior to visually cued behavioral responses and were predictive of trial-by-trial visual task performance. Finally, the task-related temporal dynamics of γ events exhibited rapid plasticity during task learning and were modality-specific. γ events in mouse V1 thus flexibly enhance visual encoding according to behavioral context.

No abstract

Modeling continuous dynamical systems from discretely sampled observations is a fundamental problem in data science. Often, such dynamics are the result of non-local processes that present an integral over time. As such, these systems are modeled with Integro-Differential Equations (IDEs); generalizations of differential equations that comprise both an integral and a differential component. For example, brain dynamics are not accurately modeled by differential equations since their behavior is non-Markovian, i.e. dynamics are in part dictated by history. Here, we introduce the Neural IDE (NIDE), a novel deep learning framework based on the theory of IDEs where integral operators are learned using neural networks. We test NIDE on several toy and brain activity datasets and demonstrate that NIDE outperforms other models. These tasks include time extrapolation as well as predicting dynamics from unseen initial conditions, which we test on whole-cortex activity recordings in freely behaving mice. Further, we show that NIDE can decompose dynamics into their Markovian and non-Markovian constituents, via the learned integral operator, which we test on fMRI brain activity recordings of people on ketamine. Finally, the integrand of the integral operator provides a latent space that gives insight into the underlying dynamics, which we demonstrate on wide-field brain imaging recordings. Altogether, NIDE is a novel approach that enables modeling of complex non-local dynamics with neural networks.