Ph.D. student. at the University of Hong Kong, advised by prof. Yi Ma Graduated from IIIS Tsinghua University 2024 (Yao Class)
Central & Western District
Joined July 2023
- Tweets147
- Following188
- Followers333
- Likes121
Pinned Tweet
Can an LLM agent play Angry Birds?
Introducing Embodied Playground 🤖
410 simulated task environments that test physical, strategic, and spatial reasoning.
🌏 Website: winniechen2002.github.io/Age…
📄 Paper: arxiv.org/abs/2610.27039
💻 Code: github.com/Winniechen2002/Em…
Feng Chen retweeted
How to do scaling research is a tough question in resource-constrained places, i.e., universities. Drawing from experience co-managing my group’s cluster and architecture project Nestor, I wrote down some thoughts and notes for junior researchers in our group, and hopefully they can be interesting to the public.
I break down the blog into 3 parts,
- a basic intro on what a scaling ladder is and why we need it, with many inspirations from @wen_kaiyue 's course.
- bitter lessons on actually landing it in academic research
- checklist before crying for chips
Some words might be spicy cuz its primary focus is internal sharing.
Fully handwritten, Opus4.6 fixed grammar, enjoy!
tianzhechu.com/blogs/scaling…
Feng Chen retweeted
Introducing Inkling 2 Small. Scoring 31 on the AA Intelligence Index, it's a big jump over its predecessor.
Inkling 2 Small is available for fine-tuning and inference on Tinker today, with full weight release planned soon.
thinkingmachines.ai/news/ink…
Congrats!
Don't trust demos. Try it yourself.
Introducing RoboAPI, powered by Poke-Aurora, our world-action model.
Give compatible robots natural-language instructions through a simple API. No task-specific training.
More information: pokebot.com/ai
In the field of language, it is still feasible, very important and valuable to explore the model structure, because it has the best evaluation and the most complete open source data in all fields. Let's take a look at Tianzhe's work. Data is the upper limit of model ability, and model structure is the key to efficiency.
You’re is always super nice!
Congrats to you and enjoy your interesting industry work!
Life update: I recently graduated from CMU RI, and last year I went through the academic faculty job market in robot learning. I was fortunate to receive a lot of help and support from many people throughout the process.
Compared to industry job applications, there are relatively few resources available online about the faculty application process. So I decided to open-source all of my application materials from last year here:
docs.google.com/document/d/1…
I hope these materials can be helpful to others who are considering pursuing an academic career!
Feng Chen retweeted
🚀T-WLA: Visual memory for fast touch
🖐️Memory mechanism connects vision and touch
⚡6× T-Rex’s decision rate (A6000 Pro)
📉30% lower finger-joint error
🤖2× T-ReX’s average success rate across 4 real-robot tasks
🎬 arxiv.org/pdf/2610.28814
@MIT @MIT_CSAIL @medialab @Harvard
Feng Chen retweeted
This is a platform to evaluate how well current frontier foundation models perform on tasks that require spatial and physical intelligence. More models and tasks will continue to be added and evaluated here: winniechen2002.github.io/Age…
Can an LLM agent play Angry Birds?
Introducing Embodied Playground 🤖
410 simulated task environments that test physical, strategic, and spatial reasoning.
🌏 Website: winniechen2002.github.io/Age…
📄 Paper: arxiv.org/abs/2610.27039
💻 Code: github.com/Winniechen2002/Em…
Feng Chen retweeted
Round of applause for Embodied Playground pushing the frontier forward with 410 simulated tasks challenging AI agents across physics, strategy, and spatial reasoning.
Great work by the team! 🚀
Can an LLM agent play Angry Birds?
Introducing Embodied Playground 🤖
410 simulated task environments that test physical, strategic, and spatial reasoning.
🌏 Website: winniechen2002.github.io/Age…
📄 Paper: arxiv.org/abs/2610.27039
💻 Code: github.com/Winniechen2002/Em…
Explore the tasks, watch the rollouts, and test your agent.
We welcome feedback, new agent evaluations, and task contributions.
Website: winniechen2002.github.io/Age…
Code: github.com/Winniechen2002/Em…
Arxiv: arxiv.org/abs/2610.27039
Thanks @PeiZhou01 @Lixuan_thu @liphilsun @YaoYunchao for collaborating and contributing.
Thanks @YiMaTweets @dingmyu @alex_kai2020 for advising.
We will keep updating the benchmark and evaluation results!