Thespian: Multi-Character Text Role-Playing Game Agents
This addresses the problem of developing more versatile AI agents for open-ended text-based games, though it appears incremental as it builds on existing agent frameworks.
The paper tackles the challenge of creating agents that can play multiple characters in text role-playing games by introducing a thespian agent framework with soft prompts and an attention mechanism for few-shot learning, showing it outperforms state-of-the-art methods in multi-character and few-shot learning.
Text-adventure games and text role-playing games are grand challenges for reinforcement learning game playing agents. Text role-playing games are open-ended environments where an agent must faithfully play a particular character. We consider the distinction between characters and actors, where an actor agent has the ability to play multiple characters. We present a framework we call a thespian agent that can learn to emulate multiple characters along with a soft prompt that can be used to direct it as to which character to play at any time. We further describe an attention mechanism that allows the agent to learn new characters that are based on previously learned characters in a few-shot fashion. We show that our agent outperforms the state of the art agent framework in multi-character learning and few-shot learning.