SR @GoogleDeepMind | PhD student @TUDarmstadt | Prev. @MIT, @Columbia. RL and Robot Learning.
Joined June 2020
- Tweets38
- Following176
- Followers313
- Likes149
📈 55% → 71.9% human-assessed success across 5 independent runs.
In a representative 6-hour run, FIND required only 30 scene-recovery interventions.
Hours of continuous real-world self-improvement, with minimal human intervention.
Work co-led with Yuan Fang, together with Haolei Tong and @liu_puze, and supervised by @GeorgiaChal.
🌐 Website: fangyzzz.github.io/FIND.gith…
💻 GitHub: github.com/FangYzzz/FIND
📄 Paper: arxiv.org/abs/2609.32069
📈 55% → 71.9% human-assessed success across 5 independent runs.
In a representative 6-hour run, FIND required only 30 scene-recovery interventions.
Hours of continuous real-world self-improvement, with minimal human intervention. 👇
Work co-led with Yuan Fang, together with Haolei Tong and @liu_puze , and supervised by @GeorgiaChal.
🌐 Website: fangyzzz.github.io/FIND.gith…
💻 GitHub: github.com/FangYzzz/FIND
📄 Paper: arxiv.org/abs/2609.32069
Built with @yjin_1118 @XiaoYangLiu10 @liu_puze @FatAndFurious42 @CarloDeramo @GeorgiaChal
🌐 Website: supersglzc.github.io/harbor-…
💻 GitHub: github.com/supersglzc/harbor…
📄 Paper: arxiv.org/abs/2606.08610
I’m currently a SR at Google DeepMind in London — Feel free to reach out!
We’ve tested HARBOR across diverse robot learning tasks, simulators, and embodiments — from tabletop manipulation to humanoid locomotion. #Robotics #ReinforcementLearning #AIAgents #ClaudeCode
Here are some of the policies autonomously built and trained by HARBOR 👇
Zechu Li retweeted
🚀 Hiring PhDs & Postdocs in Structured Robot Learning & Embodied AI @TUDarmstadt (PEARL Lab)
🤖 We study how structure in the robot–environment system can be exploited to learn robust, adaptive, and generalizable behaviors, beyond black-box policies
🔬 Topics:
• Grounding (language → perception → action)
• Structured world models
• VLA + control + memory
• RL (credit assignment, offline→online)
• Whole-body & bimanual mobile manipulation
🇪🇺 New EU Lighthouse Project on Generative AI for Robotics + ERC StG SIREN
👉 Full call: tinyurl.com/pearlgenai
Please repost 🙏
#RobotLearning #EmbodiedAI #Robotics #MachineLearning #PhDPositions #Postdoc
Unlike humans who favor a dominant hand for fine dexterous skills, robots should execute ambidextrous manipulation with equal proficiency.
Introducing SYMDEX, our algorithm for Ambidextrous Bimanual Manipulation leveraging robot’s inherent bilateral symmetry as an inductive bias.
🧩Web: supersglzc.github.io/project…,
📷Paper: arxiv.org/abs/2505.05287.
It is a pleasure to work with my amazing collaborators Yufeng Jin, @OrdonezApraez and supervisors @GeorgiaChal, @liu_puze, and Claudio Semini.
Our method significantly outperforms other diffusion-based methods on high-dimensional control benchmarks, and is also competitive with state-of-the-art non-diffusion based RL methods while requiring fewer algorithmic design choices and smaller update-to-data ratios!
I am happy to share that our work ‘DIME: Diffusion-Based Maximum Entropy Reinforcement Learning’ has been accepted to ICML 2025. Many thanks to my colleagues and collaborators @softraeh, @DenBless94, @BruceGeLi, @DPalenicek, @Jan_R_Peters, @GeorgiaChal,@geri_neumann
Zechu Li retweeted
Meet our first general-purpose robot at @DexmateAI
dexmate.ai/vega
Adjustable height from 0.66m to 2.2m: compact enough for an SUV, tall enough to reach those impossible high shelves. Powerful dual arms (15lbs payload each) and omni-directional mobility for ultimate versatility.
More videos showcasing Vega in action coming this week! Stay tuned to see this robot partner.
Introducing Vega: 🤖 @DexmateAI's newest robot that makes complex manipulation tasks simple. ✨ A step closer to intelligence and automation. 🚀
🎥 Watch now: youtu.be/PecqfiJNwQI
#Robotics #AI #Automation