LARP allows game characters to talk to players like real people, while being able to understand complex situations in the game and remember past interactions. and respond reasonably based on this information.
It can make the behavior of the game characters more realistic and deep, thereby enhancing the player’s gaming experience.
LARP works by using advanced cognitive architecture and environmental interaction modules, combined with post-processing methods, to enable AI agents in games to interact with players in a more realistic and in-depth way, thereby enhancing the overall game experience.
Cognitive architecture: Capable of mimicking human cognitive processes, including memory processing and decision-making.
Environmental Interaction Module: Interact with the game environment to learn and adapt to environmental changes.
Key features:
- More Natural Dialogue: It allows the characters in the game to talk to you in a more natural and fluid way. You can ask them questions in natural language, and they can answer in a human-like way.
- Characters have their own memories and personalities: These characters not only remember your previous interactions, but also react according to their “personality”. For example, a friendly businessman may remember the last time you helped him, while a cunning thief might try to deceive you.
- Richer Gaming Experience: These agents are able to react to events occurring in the game, providing a more diverse gaming experience. For example, if a fight takes place in the game, these characters may comment or help with it.
- Deeper Roleplay: As game characters offer deeper and varied interactions, you, as a player, can immerse yourself more deeply in roleplay and experience a richer and more realistic game world.
Project address: https://miao-ai-lab.github.io/LARP/
Paper: https://arxiv.org/abs/2312.17653
GitHub：https://github.com/MiAO-AI-Lab/LARP