Skip to content
View wadeKeith's full-sized avatar
85 followers · 15 following

Achievements

Achievement: YOLOAchievement: Pair Extraordinairex2Achievement: Starstruckx3Achievement: Pull SharkAchievement: Quickdraw

Achievements

Achievement: YOLOAchievement: Pair Extraordinairex2Achievement: Starstruckx3Achievement: Pull SharkAchievement: Quickdraw

Highlights

  • Pro

Block or report wadeKeith

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.

Content in all repositories owned by your account will be closed.
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
wadeKeith/README.md

Cheng Yin — embodied AI, VLA, and robot learning research profile

Website · Google Scholar · ORCID · Hugging Face · ModelScope · Email

I am a Ph.D. researcher at HUST, undertaking joint research training with Tsinghua THUNLP and Beijing Zhongguancun Academy. My work focuses on embodied AI, vision-language-action models, robot learning, reasoning-oriented control, and the research systems that make these models easier to train and evaluate.

Selected Research

DeepThinkVLA

FIRST-AUTHOR RESEARCH

DeepThinkVLA studies when explicit reasoning improves robot action. It reports a 97.0% average success rate on LIBERO and a 79.0% zero-shot success rate on LIBERO-Plus.

Paper · Code · Models & data

Public Contributions

Project Relationship Focus
OpenBMB/SimpleNav PUBLIC CONTRIBUTOR A unified, reproducible framework for navigation VLA research.
OpenBMB/MiniCPM-Robot PUBLIC CONTRIBUTOR An on-device embodied model family for robot manipulation and target tracking.
starVLA/starVLA PUBLIC CONTRIBUTOR A modular platform for developing and evaluating VLA systems.

Open Work & Resources

Project Relationship What it provides
Awesome-Embodied-AI MAINTAINED PUBLIC PROJECT A curated map of embodied-AI papers, datasets, simulators, toolkits, and safety resources.
autoresearch-qwen MAINTAINED PUBLIC PROJECT An autonomous research loop for improving Qwen3-VL on the public DocVQA benchmark.
Large Language and Generative Foundation Models MAINTAINED PUBLIC PROJECT A bilingual, publicly released book manuscript on large language and generative foundation models.

GitHub Activity

Public projects Stars · owned + contributed repos Contributions · last 12 months
25 579,146 1,023

Public GitHub data · repository stars include owned and verified contributed projects · refreshed 2026-09-02 04:44 UTC

Pinned Loading

  1. starVLA/starVLA starVLA/starVLA Public

    StarVLA: A Lego-like Codebase for Vision-Language-Action Model Developing

    Python 3.6k 471

  2. OpenBMB/DeepThinkVLA OpenBMB/DeepThinkVLA Public

    DeepThinkVLA: Enhancing Reasoning Capability of Vision-Language-Action Models

    Python 529 51

  3. OpenBMB/MiniCPM-Robot OpenBMB/MiniCPM-Robot Public

    A Smarter and Faster On-Device AI Brain for Robots

    Python 322 29

  4. Awesome-Embodied-AI Awesome-Embodied-AI Public

    Curated embodied AI list: surveys, VLA models, datasets, simulators, humanoids, robot learning, and safety resources.

    Python 240 13

  5. autoresearch-qwen autoresearch-qwen Public

    Autonomous Qwen3-VL training-code research on the official DocVQA benchmark. main: NVIDIA multi-GPU, mlx: Apple Silicon/MPS.

    Python 212 33

  6. OpenBMB/SimpleNav OpenBMB/SimpleNav Public

    SimpleNav: A simple, unified, reproducible, and extensible framework for navigation VLA research.

    Python 77 1