Skip to content
View lykyjxs's full-sized avatar
🎯
Focusing
🎯
Focusing
  • Peking University
  • Beijing China
  • 10:43 (UTC +08:00)

Block or report lykyjxs

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
lykyjxs/README.md

alt text


Brief Introduction

  • 👋 Hi, I'm YanKo

  • 📖 My research interests include Perception for Autonomous Driving🚗 and Embodied AI🤖

  • ✏️ The field of AI is moving incredibly fast. Continuous learning is key to navigating this shift. Welcome to my web, where I share my learning notes and projects. Let's grow together!🤝

  • 📫 You can get in touch with me by email at lykyjxs@163.com


Learning Notes

Models:

  • Transformer:Architecture breakdown of the backbone behind and source code interpretation
  • CLIP:A Vision-Language Model connecting text and images. Includes theoretical analysis and source code interpretation
  • OpenVLA-OFT:Vision-Language-Action model fine-tuning via Parameter-Efficient methods. Covers OFT ideas and results
  • Pi0:Physical Intelligence's A generalist robot foundation model based on Flow Matching.
  • Pi0-FAST:Introduces Frequency-space Action Sequence Tokenization. Optimized for high-frequency control and dexterous manipulation using DCT compression.
  • Pi0.5:Introduces multimodal web data with subtask reasoning to enable "out-of-the-box" generalization in novel home environments.
  • Pi*0.6:Uses Reinforcement Learning and value functions to self-correct and master complex long-time tasks.

AI Basic:

  • DeepLearning:Notes from Andrew Ng's Deep Learning Specialization
  • Python:Notes from Harvard's CS50P course
  • CUDA:A brief intro to CUDA programming from School of Software and Microelectronics, PKU
  • LoRA:Intuitive understanding of Low-Rank Adaptation for large model fine-tuning
  • PPO:Intuitive understanding of Proximal Policy Optimization in Reinforcement Learning, covering theoretical proofs and codebase walkthroughs
  • Flow Matching:Intuitive understanding and derivation of Flow Matching

ROS/SLAM/3D Vision:

  • VINS-Fusion:Visual-Inertial-GNSS SLAM. A deep dive into ROS system and multi-sensor fusion, covering theoretical proofs and codebase walkthroughs

Other Tools:

  • Missing-Semester:Notes from MIT course "The Missing Semester of Your CS Education," covering essential tools including Vim, Git, Shell scripting, Tmux, and pdb

Popular repositories Loading

  1. lykyjxs lykyjxs Public

  2. RLinf RLinf Public

    Forked from RLinf/RLinf

    RLinf: Reinforcement Learning Infrastructure for Embodied and Agentic AI

    Python

  3. WMPO WMPO Public

    Forked from WM-PO/WMPO

    Official Implementation of Paper: WMPO: World Model-based Policy Optimization for Vision-Language-Action Models

    Python

  4. RoboTwin RoboTwin Public

    Forked from RoboTwin-Platform/RoboTwin

    RoboTwin 2.0 Offical Repo

    Python