徐尔瀚，伦敦政治经济学院 (LSE)统计系在读一年级博士，师从史成春教授。主要研究方向包括强化学习，大语言模型的微调与优化。目前主要的研究方向为统计学方法与大预言模型的交叉应用。

Reinforcement learning focuses on rewarding desired AI actions and punishing undesired ones. Common RL algorithms include State-action-reward-state-action, Q-learning, and Deep-Q networks. RL ...

Scientists at UCL, Google DeepMind and Intrinsic have developed a powerful new AI algorithm that enables large sets of ...

WiMi's deep reinforcement learning-based task scheduling algorithm in cloud computing includes state representation, action selection, reward function and training and optimization of the algorithm.

Deep reinforcement learning leverages the learning capacity of deep neural networks to tackle problems that were too complex for classic RL techniques.

Deep reinforcement learning has helped solve very complicated challenges and will continue to be an important interest for the AI community.

Neuroscientist Daeyeol Lee discusses different modes of reinforcement learning in humans, animals, and AI, and future directions of research.

Researchers propose a method that allows reinforcement learning algorithms to accumulate knowledge while erring on the side of caution.