Decision Transformer: Reinforcement Learning via Sequence Modeling
A framework that abstracts RL as a sequence modeling problem.
Qx
Isaac Lab安装使用
Isaac Lab is unified and modular framework for robot learning that aims to simplify common workflows in robotics research.
DeepMind强化学习综述
1. Introduction 折扣因子 「强化学习交互过程存在终止风险」: 折扣因子$\gamma$的理解： 它本质是对 “交互能否持续” 的信念量化——交互持续到下一步的概率为$\gamma$，而非单纯的 “未来收益打折”（这是其表象） In general, the discount factor reflects the assumption that there i...
Robotics Transformer-2
1. 总结
Robotics Transformer-1
1. 总结
最大熵强化学习：SAC和Soft Q-learning
介绍最大熵强化学习基础理论和两个主要算法soft actor-critic和soft Q-learning
强化学习之策略梯度算法
Policy Gradient Method
PPO算法：Proximal Policy Optimization
一、引言 PPO的特点: PPO 采用了代理目标函数（surrogate object function），一个样本可以经历多个 epoch 的小批量更新，而常规的 Policy Gradient 算法每个数据样本只进行一次梯度更新。它具备 TRPO 的优点，同时更简单，样本复杂度也更优。 其他算法的不足: Q-learning：在许多简单问题上表现不佳，且其原...
强化学习之蒙特卡洛方法
Get started with Monte Carlo Method in Reinforcement Learning You will learn the basics
Customize the Favicon
The favicons of Chirpy are placed in the directory assets/img/favicons/. You may want to replace them with your own. The following sections will guide you to create and replace the default favicons...