๐ŸŒ US-Proxy
class="logged-out env-production page-responsive page-profile" style="word-wrap: break-word;" >
Skip to content
View eust-w's full-sized avatar
:shipit:
:shipit:
14.7k followers · 68 following

Achievements

Achievement: Starstruckx2Achievement: YOLOAchievement: Pair ExtraordinaireAchievement: Pull Sharkx2Achievement: Quickdraw

Achievements

Achievement: Starstruckx2Achievement: YOLOAchievement: Pair ExtraordinaireAchievement: Pull Sharkx2Achievement: Quickdraw

Highlights

Organizations

@doocs @OSTGO

Block or report eust-w

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 donโ€™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 userโ€™s behavior. Learn more about reporting abuse.

Report abuse
eust-w/README.md
Animated VLA VLM robotics banner

Longtao cat

Longtao Wu

VLA / VLM robotics builder working on robot brain architecture, simulation data, Real2Sim pipelines, and embodied AI runtime systems.

ๆœบๅ™จไบบๅคงๅฐ่„‘ ยท ๅคšๆจกๆ€ๆ„Ÿ็Ÿฅ ยท ไปฟ็œŸๆ•ฐๆฎ ยท Real2Sim ยท ๅฏ่ง‚ๆต‹่ฟ่กŒๆ—ถ

Website ยท Hugging Face ยท X ยท ไธญๆ–‡ ยท Franรงais ยท ะ ัƒััะบะธะน ยท ุนุฑุจูŠ ยท ๆ—ฅๆœฌ่ชž ยท Portuguรชs ยท Tรผrkรงe

VLA VLM Robot brain Simulation data 

robot-runtime.console
$ boot --stack vla-vlm --mode embodied
> perception=vlm  policy=vla  sim=real2sim
> brain=planner+controller  data=observable
> status=online  latency=adaptive  loop=closed

Focus

I build the stack behind robot intelligence: multimodal perception, VLA policy learning, robot big brain / small brain architecture, simulation data engines, Real2Sim assets, and runtime infrastructure that makes behavior inspectable.

Embodied AI stack summary

Moving System Map

flowchart LR
  A[World Data] --> B[Simulation Engine]
  B --> C[Embodied Dataset]
  C --> D[VLM World Model]
  D --> E[VLA Policy]
  E --> F[Robot Big Brain]
  F --> G[Small Brain Runtime]
  G --> H[Real Robot Feedback]
  H --> A

  style A fill:#020617,stroke:#22d3ee,color:#ffffff
  style B fill:#020617,stroke:#8b5cf6,color:#ffffff
  style C fill:#020617,stroke:#22c55e,color:#ffffff
  style D fill:#020617,stroke:#f59e0b,color:#ffffff
  style E fill:#020617,stroke:#fb7185,color:#ffffff
  style F fill:#020617,stroke:#38bdf8,color:#ffffff
  style G fill:#020617,stroke:#a3e635,color:#ffffff
  style H fill:#020617,stroke:#facc15,color:#ffffff
Loading

What Is Running In My Head

Layer Direction
Robot big brain multimodal planning, instruction grounding, memory, tool use
Robot small brain motion/runtime orchestration, controller adapters, execution feedback
VLA / VLM scene semantics, action grounding, policy learning, evaluation
Simulation data synthetic scenes, Real2Sim assets, domain randomization, dataset QA
AI infrastructure agents, code automation, model routing, workflow verification

Longtao Wu personal robotics style card

Selected Work Surface

My public repositories include robotics-adjacent AI infrastructure, developer tools, model routing, code review automation, knowledge workflows, and simulation/product systems. I care about systems that can be observed, debugged, reproduced, and improved instead of only looking impressive in a demo.

Area Project What it does
AI agents kakashi Codex-powered system for searching GitHub capabilities, planning repository fusion, executing changes, and verifying the result.
AI tooling ai_code_reviewer LLM-based code review automation for GitHub, GitLab, and Gitea, with multi-model support.
Model routing openai-chat-switch Go package for chat embeddings and model/chat switching workflows.
Learning systems little_language_model Small language-model experiments and implementation notes.
Developer tools esh Cross-platform SSH connection manager with encrypted credentials and cluster command execution.
Infrastructure qcow2file Builds qcow2 VM images from Dockerfile-like recipes.
Knowledge workflow obsidian-image-auto-upload Obsidian plugin for automatically uploading pasted or dropped images to external storage.

Tech Surface

Python PyTorch ROS NVIDIA Docker Kubernetes TypeScript Three.js

GitHub Signal

GitHub stats Top languages

Contribution streak

Animated coding desk at sunset

Longtao cat
VLA / VLM / Robotics / Simulation Data / Embodied AI

Pinned Loading

  1. weeklyCodingTime weeklyCodingTime
    1 
    Python     40 hrs 42 mins โ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–‰โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘  42.7%
    2 
    JSON       12 hrs 28 mins โ–ˆโ–ˆโ–‹โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘  13.1%
    3 
    Markdown   12 hrs 20 mins โ–ˆโ–ˆโ–‹โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘  12.9%
    4 
    Bash       12 hrs 7 mins  โ–ˆโ–ˆโ–‹โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘  12.7%
    5 
    Other      4 hrs 13 mins  โ–‰โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘   4.4%