
Ph.D. student at UPenn @GRASPlab | Prev Master student @UCSDJacobs, Undergrad @sjtu1896 | Working on Embodied AI & Robot Learning.
Philadelphia, PA
Joined May 2022
- Tweets129
- Following491
- Followers439
- Likes364
Geng Chen retweeted
RoboSkills agent-as-policy is now in the real world! ๐ค
- Hard coordination: twisting the lid off a bottle.
- Precise action: picking a slim pen and stacking, with no crashes.
- Cross embodiment: one workflow for both grippers and hands!
code: github.com/openrobokit/robosโฆ
Geng Chen retweeted
Our G1 robot found its way to the pantry, picked up a Pringles can, and handed it over. No pre-built map. No training for this task.
Behind it is a coding agent, writing, running, and fixing the robot's programs as it goes. And it self-evolves: what works is saved as code for the next task, no retraining needed. This is ASENA.
Joint work with @Isabella__Liu, @edmund_bu, Johan Bjorck, @yin_hongxu, @zhengyiluo, @jankautz, @DrJimFan, @yukez, and @Sifei30488L. ๐งต๐
Geng Chen retweeted
Can we scale robot learning without scaling human labor? ๐
Demonstrations, resets, failure recovery, constant supervisionโcollecting robot data still takes substantial human effort.
๐ Three keys let agents take on more of the work:
1. Real2Sim2Real: Agents explore and learn in digital twins, where trial and error is safer and cheaper.
2. Autonomous Recovery: Robots learn to recover from unexpected failures instead of waiting for human resets.
3. Human-on-Demand Learning: Multiple robots run in parallel, and humans step in only when autonomous recovery falls short.
๐ฅ Recova brings these together to get more robot learning out of every hour of human effort.
research.nvidia.com/labs/geaโฆ
Joint work with @anjjei, Johan Bjorck, @ZhidingYu, @yin_hongxu, @jankautz, @DrJimFan, @yukez, and @Sifei30488L.
๐ Here's how it works ๐งต
Geng Chen retweeted
When we started HARMONY six months ago, we were curious how far agentic approaches could take 3D reconstruction. The field has advanced quickly, and we see strong potential in combining high-level reasoning with geometric optimization. Weโre releasing an open-source framework competitive with Astra in reconstruction quality, together with the HARMONY300 benchmark. Proud of @sofine_sun and @chenwangcw!
Excited to share our new ๐ผ๐ฝ๐ฒ๐ป-๐๐ผ๐๐ฟ๐ฐ๐ฒ work, ๐๐๐ฅ๐ ๐ข๐ก๐ฌ, a framework for reconstructing compositional 3D scenes from single indoor images.
Much like a 3D artist, our VLM agent builds the scene in a ๐ต๐ถ๐ฒ๐ฟ๐ฎ๐ฟ๐ฐ๐ต๐ถ๐ฐ๐ฎ๐น ๐ผ๐ฟ๐ฑ๐ฒ๐ฟ: first the room structure and camera, then wall-mounted objects, free-standing furniture, and finally dependent decorations. After each stage, it compares its render against the input image, reflects on what went wrong, and iteratively corrects the scene to prevent errors from accumulating.
Even with older VLM backbones, our framework remains ๐ฐ๐ผ๐บ๐ฝ๐ฒ๐๐ถ๐๐ถ๐๐ฒ ๐๐ถ๐๐ต ๐๐ฃ๐ง-๐ฒ ๐๐๐๐ฟ๐ฎ ๐ถ๐ป ๐๐ถ๐๐๐ฎ๐น ๐พ๐๐ฎ๐น๐ถ๐๐, ๐๐ต๐ถ๐น๐ฒ ๐ฎ๐ฐ๐ต๐ถ๐ฒ๐๐ถ๐ป๐ด ๐ฏ๐ฒ๐๐๐ฒ๐ฟ ๐ด๐ฒ๐ผ๐บ๐ฒ๐๐ฟ๐ถ๐ฐ ๐ฎ๐น๐ถ๐ด๐ป๐บ๐ฒ๐ป๐ for single-image 3D scene reconstruction.
We also release ๐๐๐ฅ๐ ๐ข๐ก๐ฌ๐ฏ๐ฌ๐ฌ, a benchmark for single-image 3D scene reconstruction with 300 scenes across three difficulty tiers: easy, medium, and complicated.
๐ Project: cwchenwang.github.io/harmony
๐ Leaderboard: cwchenwang.github.io/harmonyโฆ
๐ค Data: huggingface.co/datasets/Shufโฆ
๐ Paper: arxiv.org/abs/2609.26793
Geng Chen retweeted
Children learn to walk by falling. Robots fail tooโbut we usually throw those failures away.
Yet failed rollouts tell us something successful demonstrations cannot: what bad actions actually lead to.
So, can robots learn from their failures too?
Introducing FACT: Failure-Aware Causal Training for World-Action Models.
๐website: fact-wam.github.io/
๐arxiv: arxiv.org/abs/2608.10232
๐ป code: github.com/Bariona/FACT
Geng Chen retweeted
Human demos contain rich hand-object skills. The challenge is the embodiment gap. Human and robot hands differ in shape, joints, and contacts.
Introducing ConTrack, our latest work that turns contact-rich human hand-object demos into robot motions.
Geng Chen retweeted
Ever want to have a single policy to control diverse robots as well as different dexterous hands, or to observe the emergent behavior under cross embodiment training?
Introducing our #CVPR2026 paper XL-VLA, Cross-Hand Latent Representation for Vision-Language-Action Models.
Geng Chen retweeted
Learning from vast video data allows point-flow-based planners to create generalizable task plans, guiding robots via future keypoint trajectories.
But how can we ensure low-level execution doesn't become the system's bottleneck?
Introducing our #ICLR2026 paper, HinFlow: Translating Flow to Policy via Hindsight Online Imitation
Geng Chen retweeted
"Cross-embodiment" is a sign of generalization. Weโve seen huge progress in manipulation and navigation โ but what about humanoid whole-body control? Can ONE policy control multiple different humanoids?
Meet our #ICRA2026 work ๐ฆ
EAGLE: Embodiment-Aware Generalist Specialist Distillation for Unified Humanoid Whole-Body Control.
Instead of brute-force URDF / morphology domain randomization, we iteratively distill specialists into one generalist. We also find that embodiment-aware representations matter for policy learning.
๐ website: eagle-wbc.github.io/
๐ arXiv: arxiv.org/abs/2602.02960
Geng Chen retweeted
VLA models are boomingโbut one fundamental question is yet seldom answered:
How does the choice of the base VLM affect VLA performanceโ๏ธ
We run a large-scale systematic study to answer this, in collaboration with @Alibaba_Qwen
cladernyjorn.github.io/VLM4Vโฆ
Geng Chen retweeted
How far can we push the limit of in-hand manipulation dexterity?
Introducing our work on motion capture: DexterCap & DexterHand !
DexterCap: A high-fidelity motion capture system for intricate in-hand manipulation motion.
DexterHand: A dataset featuring true in-hand dexterity, reorientation, finger gaiting, and even manipulating a Rubik's Cube like a speedcuber ! ๐งฉ
- Project Page: pku-mocca.github.io/Dextercaโฆ
- Experience it via our online interactive visualization: lyt0112.com/projects/DexterCโฆ
#Animation #CharacterAnimation #MotionCapture #Graphics #EmbodiedAI #DexterousManipulation
Geng Chen retweeted
What makes robot hands dexterous? ๐ค๐๏ธ
We generate robot hands + control, sim-to-real, in under 24 hours.
Paper: Cross-Embodied Co-Design for Dexterous Hands
๐ฅ Rapid evaluation w/ cross-embodied policies
๐ฆพ Open-source modular hand platform
๐ก Automated full-hand generation
Geng Chen retweeted
Meet ACE-F โ a novel, foldable teleoperation platform for collecting high-quality robot demonstration data across robot embodiments.
Using a specialized soft-controller pipeline, we interpret end-effector positional deviations as virtual force signals to provide the user with force feedback, without requiring expensive sensors! ACE-F simplifies control for a diverse array of robot platforms, making complex tasks that require dexterous manipulation highly intuitive.
Check out the project website here!
acefoldable.github.io/
Geng Chen retweeted
How do you teach a robot to do something it has never seen before? ๐ค
With human data.
Our new Human0 model is co-trained on human and humanoid data. It allows the robot to understand a novel language command and execute it perfectly in the wild without prior practice.
Real-world success rate: ๏ฝ100%.
Watch it happen ๐
A large human behavior model.
Introducing In-N-On, our latest findings in scaling egocentric data for humanoids.
1. Pre-training and post-training with human data
2. 1,000+ hours of in-the-wild data and 20+ hours of on-task data with accurate action labels
Website: xiongyicai.github.io/In-N-Onโฆ
Arxiv: arxiv.org/abs/2511.15704
By simply scaling data, our robot can follow novel language instruction. Check out the ๐งต
Geng Chen retweeted
A large human behavior model.
Introducing In-N-On, our latest findings in scaling egocentric data for humanoids.
1. Pre-training and post-training with human data
2. 1,000+ hours of in-the-wild data and 20+ hours of on-task data with accurate action labels
Website: xiongyicai.github.io/In-N-Onโฆ
Arxiv: arxiv.org/abs/2511.15704
By simply scaling data, our robot can follow novel language instruction. Check out the ๐งต
Geng Chen retweeted
Meet BFM-Zero: A Promptable Humanoid Behavioral Foundation Model w/ Unsupervised RL๐ lecar-lab.github.io/BFM-Zeroโฆ
๐งฉONE latent space for ALL tasks
โกZero-shot goal reaching, tracking, and reward optimization (any reward at test time), from ONE policy
๐คNatural recovery & transition
Geng Chen retweeted
Teaching robots sub-millimeter precision through spatially-grounded touch
We introduce SaTA which anchors tactile to hand kinematics, enabling learning-based policies to match model-based precision.
arxiv.org/abs/2510.14647
Geng Chen retweeted
Ever want to enjoy all the privileged information in sim while seamlessly transferring to the real world? How can we correct policy mistakes after deployment?
๐Introducing GSWorld, a real2sim2real photo-realistic simulator with interaction physics with fully open-sourced code.

















