Ann Kennedy
Co-PI (Core Leadership)
Northwestern University
Ann Kennedy, PhD, is a theoretical neuroscientist at Scripps Research in San Diego. She is broadly interested in the neural control of complex and naturalistic behavior, and her research has focused on the use of methods from machine learning, dynamical systems, and control theory to characterize the structure of animal behavior and its control by the brain. She completed postdoctoral training with Dr. David Anderson at California Institute of Technology, where she modeled hypothalamic control of social and fear behaviors. Prior to that, she pursued her PhD at Columbia University with Dr. Larry Abbott, where she modelled information representation and learning in cerebellum-like structures.
Recent ASAP Preprints & Published Papers
Neural Heterogeneity Controls Computations in Spiking Neural Networks
Significance Neurons are the basic information-encoding units in the brain. In contrast to information-encoding units in a computer, neurons are heterogeneous, i.e., they differ substantially in their electrophysiological properties. How does the brain make use of this heterogeneous substrate to carry out its function of processing information and generating adaptive behavior? We analyze a mathematical model of networks of heterogeneous spiking neurons and show that neural heterogeneity provides a previously unconsidered means of controlling computational properties of neural circuits. We furthermore uncover different capacities of inhibitory vs. excitatory heterogeneity to regulate the gating of signals vs. the encoding and decoding of information, respectively. Our results reveal how a mostly overlooked property of the brain—neural heterogeneity—allows for the emergence of computationally specialized networks. Abstract The brain is composed of complex networks of interacting neurons that express considerable heterogeneity in their physiology and spiking characteristics. How does this neural heterogeneity influence macroscopic neural dynamics, and how might it contribute to neural computation? In this work, we use a mean-field model to investigate computation in heterogeneous neural networks, by studying how the heterogeneity of cell spiking thresholds affects three key computational functions of a neural population: the gating, encoding, and decoding of neural signals. Our results suggest that heterogeneity serves different computational functions in different cell types. In inhibitory interneurons, varying the degree of spike threshold heterogeneity allows them to gate the propagation of neural signals in a reciprocally coupled excitatory population. Whereas homogeneous interneurons impose synchronized dynamics that narrow the dynamic repertoire of the excitatory neurons, heterogeneous interneurons act as an inhibitory offset while preserving excitatory neuron function. Spike threshold heterogeneity also controls the entrainment properties of neural networks to periodic input, thus affecting the temporal gating of synaptic inputs. Among excitatory neurons, heterogeneity increases the dimensionality of neural dynamics, improving the network’s capacity to perform decoding tasks. Conversely, homogeneous networks suffer in their capacity for function generation, but excel at encoding signals via multistable dynamic regimes. Drawing from these findings, we propose intra-cell-type heterogeneity as a mechanism for sculpting the computational properties of local circuits of excitatory and inhibitory spiking neurons,…
PyRates--A Code-Generation Tool for Dynamical Systems Modeling
We present PyRates, a code-generation tool for dynamical systems modeling applied to biological systems. Together with its extensions PyCoBi and RectiPy, PyRates provides a framework for modeling and analyzing complex biological systems via methods such as parameter sweeps, bifurcation analysis, and model fitting. We highlight the main features of this framework, with an emphasis on new features that have been introduced since the initial publication of the software, such as the extensive code generation capacities and widespread support for delay-coupled systems. Using a collection of mathematical models taken from various fields of biology, we demonstrate how PyRates enables analysis of the behavior of complex nonlinear systems using a diverse suite of tools. This includes examples where we use PyRates to interface a bifurcation analysis tool written in Fortran, to optimize model parameters via gradient descent in PyTorch, and to serve as a model definition interface for new dynamical systems analysis tools.
