Founder & Chief Scientist of University AI. A watchful guardian for AGI.
New ideas
由 Tom Everitt,Ramana Kumar 和 Marcus Hutter 撰写
By Victoria Krakovna (DeepMind), Ramana Kumar (DeepMind), Laurent Orseau (DeepMind), Alexander Turner (Oregon State University)…
By Tom Everitt, DeepMind
Translated by Xiaohu Zhu, Founder of University AI, contact: neil@universityai.com
我们在最新的论文中,描述了一个新的推断智能体动机的方法。该方法基于影响图,这是一种图模型的类型,包含特别的决策和效用节点。图标准可以被用来确智能体观测动机和智能体干预动机
# Scalable agent alignment via reward modeling
By Pedro A. Ortega, Vishal Maini, and the DeepMind safety team