Towards Controllable Agent in MOBA Games with Generative Modeling
This work addresses the challenge of creating human-like and alignable agents in MOBA games, which is incremental as it builds on existing generative modeling approaches.
The paper tackled the problem of developing action-controllable agents in MOBA games by modeling control as an action generation process, resulting in a deep latent alignment neural network model and sampling algorithm that demonstrated efficacy in simulated and online experiments in Honor of Kings.
We propose novel methods to develop action controllable agent that behaves like a human and has the ability to align with human players in Multiplayer Online Battle Arena (MOBA) games. By modeling the control problem as an action generation process, we devise a deep latent alignment neural network model for training agent, and a corresponding sampling algorithm for controlling an agent's action. Particularly, we propose deterministic and stochastic attention implementations of the core latent alignment model. Both simulated and online experiments in the game Honor of Kings demonstrate the efficacy of the proposed methods.