OpenAI News·· 2018-10-31AI 评分37
OpenAI 提出 RND 方法，强化学习智能体首次超越人类在 Montezuma’s Revenge 的表现
Reinforcement learning with prediction-based rewards
AI 导读
OpenAI 开发了基于预测的随机网络蒸馏（RND）方法，通过好奇心驱动强化学习智能体探索环境。该方法首次在 Montezuma’s Revenge 游戏中实现超过人类平均水平的表现。
原文
We’ve developed Random Network Distillation (RND), a prediction-based method for encouraging reinforcement learning agents to explore their environments through curiosity, which for the first time exceeds average human performance on Montezuma’s Revenge.
来源：OpenAI News · openai.com