Browsing by Author "Williams, Ross"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
- Generative Agent-Based Modeling: Unveiling Social System Dynamics through Coupling Mechanistic Models with Generative Artificial IntelligenceGhaffarzadegan, Navid; Majumdar, Aritra; Williams, Ross; Hosseinichimeh, Niyousha (2024-01)We discuss the emerging new opportunity for building feedback-rich computational models of social systems using generative artificial intelligence. Referred to as Generative Agent-Based Models (GABMs), such individual-level models utilize large language models to represent human decision-making in social settings. We provide a GABM case in which human behavior can be incorporated in simulation models by coupling a mechanistic model of human interactions with a pre-trained large language model. This is achieved by introducing a simple GABM of social norm diffusion in an organization. For educational purposes, the model is intentionally kept simple. We examine a wide range of scenarios and the sensitivity of the results to several changes in the prompt. We hope the article and the model serve as a guide for building useful dynamic models of various social systems that include realistic human reasoning and decision-making.
- Traffic Calming at Virginia TechAlford, Ella; Barlow, Grant; Cummins, Nicholas; Ellsworth, Zoe; Williams, Ross (2024-05-01)Congestion on Virginia Tech’s campus is a growing problem with increased foot, bike, car, and bus traffic. This project explores the source of the congestion, when congestion is at a peak, and why. The main findings center around raw data that finds the peak class times to induce surges of cars, buses, pedestrians, bikes, and micromobility travel. The problem stems from pedestrians and multimodal forms of transport having inconsistent and inadequate signals at crosswalks, causing a backup in traffic. Not only does this decrease the level of safety, but it increases travel time. We explore how to solve this problem by looking at national and international solutions and proposing multiple solutions below.