Zoe
Hareng
Sponsor: Rishidev Chaudhuri, Ph.D. Neuro Physio & Behavior Many theories about the brain regard it to be continuously processing data, making predictions, and correcting mistakes. But with what kinds of networks and dynamics could it do that is the question we hope to answer. Networks of neurons in the brain have the potential to be exceedingly chaotic. While chaos and noise are often seen as impediments to computation, we show how chaotic activity may instead be used to support generative models in the brain. Specifically, we show that variability corresponding to the chaotic dynamics of strongly coupled recurrent neural networks can be used as a novel sampling scheme for generative models. Furthermore, we propose that the overall network gain parameter could be changed by neuromodulation to control the rate of learning and sampling in these models. As a proof-of-principle, we demonstrate the idea by adapting conventionally used generative models. While the sampling method is slow in traditional digital computing paradigms, we emphasize its applicability in the ability to implement generative architectures both in distributed biological systems and in neuromorphic hardware. Virtual Reality Shopping: How Does Self- Reported Distractibility Affect Search Behavior in VR?
Abstract profile. Full document pending author claim.
Authors:
Zoe Hareng
Date Created:
Not specified
Course Title:
Professor:
Not specified
About Paper:
Various studies have begun to investigate search behavior in virtual reality (VR) environments, leveraging the ability to manipulate settings realistically while measuring aspects of behavior. Despite the evolving landscape of VR research, a gap exists in studies capturing organic search behavior. Studies have shown that people with high distractibility, such as those diagnosed with ADHD, exhibit less efficient visual search. However, this has primarily been studied using 2D displays, so it is unclear whether these findings extend to 3-dimensional search behaviors. This study addressed this limitation by having participants explore an IKEA-like VR furniture store, followed by a surprise "shopping" search task while their eye movements and position were tracked. Subsequently, participants complete a questionnaire, including the BAARS Inattention Scale to assess trait distractibility. The central question of interest revolves around how self-reported distractibility influences facets of search behavior. Preliminary results (N=53) suggest a small correlation between higher trait inattention scores and lower search efficiency across participants. Additional analyses will assess whether distractibility scores are related to various aspects of search behavior, such as time to find targets and fixations made during search. This research contributes to our understanding of how trait distractibility influences real-world behaviors, specifically within three-dimensional environments. UC Davis 35 th Annual Undergraduate Research, Scholarship and Creative Activities Conference 113 Assessing Changes in Soil Fusarium oxysporum and Fungal Profiles in Response to Biosolarization with Date Residue and Rice Bran Amendments in a Desert Climate Krishnapriya Hari
Source:
UC Davis / Psychology / 2024
Topics:
No topics listed
Co-authors:
Zoe Hareng