Citation
McGill, Mason Benjamin (2025) Visual Systems and the Forces That Shape Them. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/y27w-m760. https://resolver.caltech.edu/CaltechTHESIS:06092025-113042101
Abstract
Vision neuroscience provides a unique opportunity to draw a correspondance between the physical world and its neural representation. But despite the amazing advances in neural recording technology that have occurred over the past two decades, we can't yet simultaneously record from more than a tiny fraction of the neurons in most of the visual systems currently being studied, which limits our ability to develop a holistic cause-and-effect understanding of how they operate. So it may make sense, as a complement to directly studying a visual system found in nature, to also study synthetic visual systems that in some way resemble it but are easier to inspect. This document describes four lines of work aimed at improving our ability to learn about biological visual systems using models optimized in ways that are analogous to the selective pressures that biological visual systems face, like the pressures to relay accurate information about the world, minimize energy consumption, and withstand perturbation. The first two of these lines of work---discussed in chapters 2 and 3---focus on expanding the space of selective forces that can be factored into optimization-guided models, and the other two---discussed in chapters 4 and 5---focus on modeling particular visual systems (in the macaque and the fruit fly, respectively). Taken together, optimization-guided modeling is shown to be a promising approach to advancing our understanding of visual processing across the animal kingdom, allowing us to leverage hypotheses about the high-level properties of visual systems to amplify the value of sparse neural data.
| Item Type: | Thesis (Dissertation (Ph.D.)) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Subject Keywords: | Vision, neuroscience, evolution, machine learning | ||||||||||||
| Degree Grantor: | California Institute of Technology | ||||||||||||
| Division: | Biology and Biological Engineering | ||||||||||||
| Major Option: | Computation and Neural Systems | ||||||||||||
| Thesis Availability: | Not set | ||||||||||||
| Research Advisor(s): |
| ||||||||||||
| Thesis Committee: |
| ||||||||||||
| Defense Date: | 25 March 2025 | ||||||||||||
| Funders: |
| ||||||||||||
| Record Number: | CaltechTHESIS:06092025-113042101 | ||||||||||||
| Persistent URL: | https://resolver.caltech.edu/CaltechTHESIS:06092025-113042101 | ||||||||||||
| DOI: | 10.7907/y27w-m760 | ||||||||||||
| Related URLs: |
| ||||||||||||
| ORCID: |
| ||||||||||||
| Default Usage Policy: | No commercial reproduction, distribution, display or performance rights in this work are provided. | ||||||||||||
| ID Code: | 17429 | ||||||||||||
| Collection: | CaltechTHESIS | ||||||||||||
| Deposited By: | Mason McGill | ||||||||||||
| Deposited On: | 09 Jun 2025 22:27 | ||||||||||||
| Last Modified: | 17 Jun 2025 17:10 |
Full text not available from this repository.
Repository Staff Only: item control page
