Task-driven Modeling of the Drosophila Head Direction Circuits

<a href="https://embed.stanford.edu/iframe/?url=https%3A%2F%2Fpurl.stanford.edu%2Fqm555fh7140" class="su-underline">Show Content</a>

Abstract/Contents

Abstract: Animals rely on internal representations of their head direction to stay oriented while navigating their environments. In the Drosophila brain, these representations are instantiated by head direction cells in a brain region called the central complex. Like the head direction circuits of other animals, the central complex exhibits dynamics that can be modeled by a class of recurrent neural networks called ring attractors. Earlier work has shown that artificial neural networks trained to perform angular velocity integration, the tracking of an agent’s heading angle from its angular velocity over time, naturally recapitulate ring attractor dynamics and other features of biological head direction circuits. In this paper, I build upon this earlier task-driven modeling work, further characterizing the dynamics and connectivity of RNNs trained to perform a modified form of angular velocity integration. Like the preceding study, I find that these trained models develop single unit response properties that mirror those of biological head direction cells, display low-dimensional ring attractor-like dynamics, and develop a connectivity pattern of short-range excitation and long-range inhibition characteristic of ring attractor networks. In addition, I also explore how these networks behave in the absence of inputs and respond to the stimulation of subsets of their units.

Type of resource	text
Date created	2021

Author	Osman, Mohammed Abdal Monium
Primary advisor	Clandinin, Thomas
Advisor	Druckmann, Shaul
Degree granting institution	Stanford University, Symbolic Systems Program

Location	https://purl.stanford.edu/qm555fh7140

Use and reproduction: User agrees that, where applicable, content will not be used to identify or to otherwise infringe the privacy or confidentiality rights of individuals. Content distributed via the Stanford Digital Repository may be subject to additional license and use restrictions applied by the depositor.
License: This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).

Preferred Citation: Osman, Mohammed Abdal Monium. (2021). Task-driven Modeling of the Drosophila Head Direction Circuits. Stanford Digital Repository. Available at: https://purl.stanford.edu/qm555fh7140

Undergraduate Honors Theses, Symbolic Systems Program, Stanford University

Loading usage metrics...