---

layout: page

permalink: /blogs/RobotLearning/index.html title: RobotLearning

具身智能补完计划

  这是一个雄心勃勃的学习计划及学习记录：利用大四学年补完机器人+AI的基础知识。之前急功近利做项目其实欠下了不少“债”，希望能持续更新，争取每天都学一点。

Python

cs231n python tutorial：基本语法、numpy✅

Robot Learning

课程：Youtube-ETHZ Oier Mees-Robot Learning (Link)

课程内容

• 1 Introduction to Robot Learning Slides Recording
  • 讨论：No paper discussion.
  • 笔记：10.17✅
• 2 Robot Control & MDPs Slides Recording
  • 讨论
    • Simple random search provides a competitive approach to RL (Mania et al., 2018)
    • Deep RL Doesn’t Work Yet (Irpan, 2018)
    • Curiosity-driven Exploration by Self-supervised Prediction (Pathak et al., 2017)
  • 嘉宾：Abishek Gupta (Prof. University of Washington) YouTube Recording
  • 笔记：
    • 10.17看课程✅
• 3 Imitation Learning Slides Recording
  • 讨论
    • Causal Confusion in IL (Den Haan et al., 2019)
    • The surprising effectiveness of representation learning for visual imitation (Pari et al. 2021)
    • Transporter Networks: Rearranging the Visual World for Robotic Manipulation (Zeng et al., 2020)
  • 嘉宾：Danfei Xu (Prof. Georgia Tech) YouTube Recording
• 4 Reinforcement Learning I Slides Recording
  • 讨论
    • Evolution Strategies as a Scalable Alternative to RL (Salimans et al., 2017)
    • Learning Synergies between Pushing and Grasping (Zeng et al., 2018)
    • Precise and Dexterous Robotic Manipulation via Human-in-the-Loop Reinforcement Learning (Luo et al., 2024)
  • 嘉宾：Aviral Kumar (Prof. Carnegie Mellon University & Google DeepMind) YouTube Recording
• 5 Reinforcement Learning II Slides Recording
  • 讨论
    • End-to-End Training of Deep Visuomotor Policies (Levine et al., 2015)
    • Eureka: Human-Level Reward Design via Coding LLMs (Ma et al., 2023)
    • Latent Plans for Task Agnostic Offline Reinforcement Learning (Rosete-Beas et al., 2022)
  • 嘉宾：Andrew Wagenmaker (Postdoc UC Berkeley) YouTube Recording
• 6 Generative Models Slides Recording
  • 讨论
    • Planning with Diffusion for Flexible Behavior Synthesis (Janner & Du et al., 2022)
    • Implicit Behavioral Cloning (Florence et al., 2021)
    • Steering Your Diffusion Policy with Latent Space RL (Wagenmaker et al., 2025)
  • 嘉宾：Cheng Chi (Co-Founder Sunday Robotics, Lead of Diffusion Policy & UMI) YouTube Recording
• 7 Sequence Modeling and Transformers Slides Recording
  • 讨论
    • Decision Transformer: RL via Sequence Modeling (Chen et al., 2021)
    • Learning Fine-Grained Bimanual Manipulation (ALOHA) (Zhao et al., 2023)
    • Humanoid Locomotion as Next Token Prediction (Radosavovic et al., 2024)
  • 嘉宾：Ted Xiao (Co-Founder Prometheus, ex-Google) YouTube Recording
• 8 World Models Slides Recording
  • 讨论
    • Learning Universal Policies via Text-Guided Video Generation (Du et al, 2023)
    • Training Agents Inside of Scalable World Models (Hafner et al., 2025)
    • World Action Models are Zero-shot Policies (Ye et al., 2026)
  • 嘉宾：Scott Reed (Principal Research Scientist NVIDIA GEAR Lab) YouTube Recording
• 9 Generalist Robot Policies Slides Recording
  • 讨论
    • Language Conditioned Imitation Learning over Unstructured Data (Lynch et al., 2021)
    • A Generalist Agent (Gato) (Reed et al., 2022)
    • π∗0.6: a VLA That Learns From Experience (Physical Intelligence, 2025)
  • 嘉宾：Quan Vuong (Co-Founder Physical Intelligence) YouTube Recording
• 10 Embodied Reasoning and Test-time Scaling Slides Recording
  • 讨论
    • In-Context Imitation Learning via Next-Token Prediction (Fu et al., 2024)
    • VOYAGER: An Open-Ended Embodied Agent with LLMs (Wang et al., 2023)
    • Training Strategies for Efficient Embodied Reasoning (Chen et al., 2025)
  • 嘉宾：Archit Sharma (Research Scientist Google DeepMind, Co-creator Gemini Deep Think series) YouTube Recording
• 11 Frontier & Open Problem Slides Recording
  • 讨论
    • A Path Towards Autonomous Machine Intelligence (LeCun, 2022)
    • The Bitter Lesson (Sutton, 2019)
    • Intelligence without Representation (Brooks, 1991)
  • 嘉宾：Lucas Beyer (Meta Superintelligence Labs) YouTube Recording
• 12 Guest Lectures
  • 嘉宾：Dieter Fox (Prof. University of Washington & Director AI2) (confirmed)

作业

• 1 Pytorch & Numpy Tutorial Code
  • ex1✅：tensor基础（声明、manipulation转换大小、索引、广播/归约、向量化：无需循环实现快速操作、einsum） -> vibe coding过了一遍，用处不大即如何用pytorch实现各类运算
  • ex2：自动微分
• 2 Robot Control & MDPs Code
• 3 Imitation Learning Code
• 5 Reinforcement Learning Code

机器人学

Youtube-Northwestern Kevin Lynch-Modern Robotics

动力学/力控/阻抗导纳

李群李代数/流形

雅可比

AI

Youtube-Stanford Fei-fei Li et el.-CS231n Deep Learning for Computer Vision (Link)

一般网络搭建、训练

transformer

Generative Model

c++