BriefGPT.xyz
Sep, 2019
无新闻外交:多代理人游戏建模
No Press Diplomacy: Modeling Multi-Agent Gameplay
HTML
PDF
Philip Paquette, Yuchen Lu, Steven Bocco, Max O. Smith, Satya Ortiz-Gagne...
TL;DR
该研究使用专家轨迹训练了一个基于神经网络的无语版外交政策模型,然后使用强化学习代理在自我对弈过程中进行了训练,两种代理表现均超过了基于规则的机器人。
Abstract
diplomacy
is a seven-player non-stochastic, non-cooperative game, where agents acquire resources through a mix of teamwork and betrayal. Reliance on trust and coordination makes
diplomacy
the first non-cooperativ
→