Neural population dynamics

Neural systems form high-dimensional representations that underpin their computational capabilities. Consequently, methods for quantifying representational similarity in neural representations have become a popular tool for identifying computational principles that are potentially shared across neural systems. These methods generally assume that neural responses are deterministic and static. However, responses of biological systems are noisy and dynamically unfold over time and these characteristics can have substantial influence on a system’s computational capabilities. We first show that exi

Neural population dynamics are shaped by many cellular, synaptic, and network properties. Not only is it important to understand how coordinated changes in circuit parameters alter neural activity, but also when dynamics are unaffected by---or invariant to---such changes. Computational modeling has revealed invariances in single neurons and small circuits that are thought to reflect their robustness against variability and perturbations. However, generalizing these insights to larger circuits in cortex and other brain areas remains challenging. A key bottleneck lies in inverse modeling of neur

…movement, but it remains unclear the neural computation that preparatory activity develops to initialize movements in dynamic sensorimotor contingency, such as interception of motion targets. In the present study, we analyzed neural population dynamics recorded from motor cortex via Utah array while two macaques performed a delayed reach toward a target either static or moving circularly. Principal component analysis (PCA) showed that the directional tuning of preparatory latent fact

Recent advances in techniques for monitoring and perturbing neural populations have greatly enhanced our ability to study circuits in the brain. In particular, two-photon holographic optogenetics now enables precise photostimulation of experimenter-specified groups of individual neurons, while simultaneous two-photon calcium imaging enables the measurement of ongoing and induced activity across the neural population. Despite the enormous space of potential photostimulation patterns and the time-consuming nature of photostimulation experiments, very little algorithmic work has been done to dete

Neural population dynamics are jointly shaped by numerous factors at the cellular, synaptic, and network levels. Mechanistic models, especially spiking neural networks, are crucial for understanding how these interacting variables influence neural activity. However, the process of adjusting model parameters to match experimental recordings, i.e., inverse modeling, is challenging, and often requires a trade-off between model complexity and tractability. Consequently, models often use fixed or manually tuned parameters, limiting their dynamic range and explanatory power while making cross-study

Artificial recurrent neural networks (RNNs) have become a customary research tool in theoretical neuroscience, being used as models of biological neural networks performing cognitive tasks. While RNNs are not generally intended as faithful, biologically realistic models of neuronal activity, there is a broad expectation that RNN dynamics may provide good models at the level of neural population dynamics. According to several authors, in both cases population activity should organize into low-dimensional manifolds encoding the relevant task variables (1). However, strong tests of this hypothesi

…synchrony and apply it to models of neural population dynamics. Conventionally, OCT requires a target trajectory. This requirement may be overly restrictive for oscillatory targets since the exact trajectory shape including amplitude and phase might not be relevant. To overcome this limitation, we introduce three alternative cost functionals to target oscillations and synchrony without specifying a reference trajectory. A controlled nonlinear dynamical system with state $x$ and control $u$ is defined by $$\dot

…emerge from the idiosyncratic neural circuitry of each individual. The overall organisation of neural circuits is preserved across individuals because of their common evolutionarily specified developmental programme. Such organisation at the circuit level may constrain neural activity, leading to low-dimensional l

Variations in cellular and network parameters shape neural dynamics and computation. Mechanistic models, such as spiking neural networks (SNNs), are instrumental for linking recordings of neural population dynamics to their underlying circuit parameters. However, SNNs lack analytical expressions and accessible gradients for parameter inference (i.e., model identification), making it difficult to systematically identify a single configuration that can reproduce experimental data, let alone the full space of data-consistent models and their uncertainty. As a result, SNNs are often reduced a prio

…have shown that high-dimensional neural population dynamics lie within low-dimensional manifolds, which may be integral to neural control of behavior. However, many of these studies have focused on neural activity recorded during experimentally driven movements with little variability, leaving open the question of whether low-dimensional neural manifolds persist during naturalistic movements. In this paper, we investigate neural population dynamics in naturalistic reaching

…models of the brain like recurrent neural networks (RNNs) often have their weights frozen after training, relying on an internal state stored in neuron activity to process temporal information. Although networks with dynamical synapses have been explored previously, often said dynamics are added to networks that also have recurrent connections and thus the short-time scale computational capa

…and the widespread adoption of neural population frameworks, there is an increased need to achieve efficient decoding of task-relevant behavioral information from neural dynamics. The most prevalent methods to uncover task-related topologies in population dynamics, often approximated as neural manifolds, are linear dimensionality reduction methods such as principal component analysis (PCA). These techniques, however, find large-variance subspaces within which to embed neural manifolds and such embeddings do not necessaril

Neural dynamics in a given brain region are generated either due to recurrent connections in the region (intrinsic dynamic), or due to temporally structured external inputs (input dynamics). These inputs can be sensory stimuli or excitations from upstream brain regions. While measuring all inputs is infeasible experimentally, measurements of sensory inputs such as task instructions or partial measurements of neural inputs into a brain region are often possible. When learning models of neural population dynamics, current learning methods can either account for measured input but not behavior, o

The physiological properties of neurons influence their population dynamics to shape neural computation and behavior. These physiological properties are diverse: the brain is composed of many genetically distinct neuronal cell types with pronounced physiological and functional differences. But neurons of the same cell type are also not identical, and exhibit heterogeneity in their physiological properties. We call this "within-type" heterogeneity, in contrast to the former, "between-type" heterogeneity. Whereas the contributions of distinct neuron types to neural network dynamics has been stud

The cost of experimental neural recordings strongly limits the stimuli or conditions one can present. Ideally, we would therefore like to know how neural responses observed in specific experiments generalize beyond them. E.g., how does the brain’s response to isolated sounds presented in the lab compare to its response to those sounds within natural language in complex social settings? We addressed this problem in fly courtship–a complex, acoustically mediated social behavior. Starting with recorded responses from diverse female neurons to isolated elements of male courtship song, we found the

It has been commonly reported across species and brain regions that neural population dynamics are low-dimensional, leading to a broad range of theories on the significance of this structure [1, 2]. However, several recent studies have reported that spontaneous activity in rodent sensory regions is high-dimensional [3], presenting a challenge to explanations which posit low-dimensional dynamics as integral to neural computations. Here we hypothesized that the observation of high-dimensional population dynamics in these datasets results from neural activity progressing through a series of low-d

Recurrent Neural Networks (RNN) are useful models to study neural computation. Several approaches are available to train RNNs on neuroscience-related tasks and reproduce neural population dynamics. However, a comprehensive understanding of how the generated dynamics emerge from the underlying connectivity is largely lacking. Previous work derived such an understanding for specific types of constrained RNNs (Mastrogiuseppe and Ostojic, 2020; Curto et al., 2019; Biswas and Fitzgerald, 2020), but if and how the resulting insights apply more generally is unclear. In this work, we study how network

According to the theory of predictive processing, expectations modulate neural activity so as to optimize the processing of sensory inputs expected in the current environment. While there is accumulating evidence that the brain indeed operates under this principle, most of the attention has been placed on mechanisms that rely on static coding properties of neurons. The potential contribution of dynamical features, such as those reflected in the evolution of neural population dynamics, has thus far been overlooked. In this talk, I will present evidence for a novel mechanism for predictive proce