Behavior Trees for Human-Agent Collaboration: Generate, Adapt, Use with Humans and LLMs
Suitable for:
Bachelor or Master projectDescription:
This thesis explores the use of behavior trees (BTs) for modeling and supporting collaboration between humans and artificial agents. The project focuses on methods for generating and adapting BTs, with large language models (LLMs) serving as tools to support these processes. Human users may guide the adaptation of BTs with the assistance of LLMs, or the project may focus on representing human behavior within BTs. Implementation will be in Python, with evaluation in cooperative scenarios involving human participants.