Embodied Cognition and Self-Concept of GPT-4 Turbo Vision through Robotics

Thumbnail
Name:
FRS_2024_46.pdf
Size:
411.0Kb
Format:
PDF
Download
Cast your vote
 
Your vote was cast
Thank you for your feedback
Author
Victor Lockwood
John Lindstedt
Bastian Tenbergen
Date Published
2024

Metadata
Show full item record
URI
http://hdl.handle.net/20.500.12648/15659
Abstract
GPT-4 Turbo Vision, a large language model (LLM), presents unique opportunities to investigate embodied cognition. We present a system that allows the LLM to be “embodied� in its environment via a robot. We conduct a mirror test and find that the LLM is able to recognize itself in the mirror.
Collections