Reasoning in VLA Models
Professor: Moayed D Daneshyari
Description: This project is about how a physical AI system, robotics system, learns to reason. The robots nowadays are trained on multimodal LLM known as Vision-Language-Action (VLA) models. In this project, we extend the VLA model by enhancing the reasoning into the model using methods such as Chain-of-Thought or other techniques.
Preferred Qualifications: Basic knowledge of python programming.
