LLM 引导强化学习提升多智能体战斗游戏 NPC 适应性
原标题：LLM-Guided Reinforcement Learning for Adaptive NPC Behavior in Multi-Agent Combat Games
AI 摘要
本研究测试了本地运行的 Mistral 7B 大语言模型（通过 Ollama）每五秒读取实时游戏状态并为共享 PPO 策略分配战术标签，以增强多智能体战斗游戏中 NPC 的适应性。实验表明，在对抗平衡型对手时，增强小队胜率从 11% 提升至 24%，但在对抗激进型对手时，模型过度偏好包围策略导致效果不佳。分析显示 83.8% 的标签选择为包围，表明该规模模型在零样本策略区分上存在局限。
正文节选
LLM-Guided Reinforcement Learning for Adaptive NPC Behavior in Multi-Agent Combat Games Abstract. Non-player characters in combat video games have long frustrated players and designers alike. Scripted enemies follow fixed patterns that experienced players learn to exploit within minutes, whilst purely rule-based systems offer no mechanism for adjusting to what a specific opponent is actually doing. Reinforcement learning offered a partial solution, producing agents that learn effective behaviour