
Research Scientist @Nvidia, PhD @UNC Chapel Hill, 3D vision, video gen, robotics, telepresence. MSCV @CarnegieMellon, ECE @UofIllinois .
Joined October 2020
- Tweets87
- Following346
- Followers68
- Likes298
Pinned Tweet
Grounded-Exo2Ego accepted to NeurIPS 2026 🎉 The single biggest limiting factor of Robotics is data. Our vision is to turn the ocean of long, in-the-wild, exocentric human videos on the internet into egocentric videos that robots can learn from.
shengze wang retweeted
Film once. Move the camera later. 🎥
Our video model reconstructs what you see and generates what you don’t. Feed it a rough 4D reconstruction, get back a dynamic world you can explore in real time!
Led by the amazing @liyuan_zz. Come see how at #NeurIPS2026!
Glad to share that World from Motion has been accepted to #NeurIPS2026! We’re releasing the code, model weights and DyCheck reproduction scripts:
Code: github.com/GradientSpaces/Wo…
Model weights: huggingface.co/liyzzz/World-…
See you in Sydney 🌏🇦🇺🐨!
#neurips2026 #NVIDIA @alextrevith
shengze wang retweeted
PingpongDuo
Humanoid table tennis now has doubles, while I'm still single.
proj: archon.tech/research/Pingpon…
abs: arxiv.org/abs/2610.20579
shengze wang retweeted
Now autoregressive video models can remember. 🧠
Introducing MosaiChunk: persistent visual memory for long-horizon autoregressive video generation🧩.
Instead of remembering everything, it remembers the right things. ✍️
1/6
shengze wang retweeted
Casual photos in, simulation-ready 3D objects out! Introducing 🐊GATOR🐊, generative + agentic 3D object reconstruction. Here, humanoids interact with GATOR-reconstructed furniture [1/n]
Website: research.nvidia.com/labs/lpr…
Paper: arxiv.org/abs/2610.11215
shengze wang retweeted
We will present DyaPlex at #NeurIPS2026, a full-duplex human-robot interaction model, which allows AI (both agents and robots) to listen, speak, perceive partner motion, and generate synchronized motion in full-duplex dyadic interaction. research.nvidia.com/labs/amr…
shengze wang retweeted
Joint work with @NVIDIAAI and @hkust with @Kumapower1 @swookpark @_TianyeLi @amritamaz Christian Jacobsen @mct1224 @virtualitaet @rjrshr Ka Chun Cheung Simon See @shalinidemello
shengze wang retweeted
Our team at NVIDIA @NVIDIAAI is hiring a research intern for 2027. Please send me an email if you are interested in working on technology to enable Interactive Physical AI (research.nvidia.com/labs/amr…), especially human #robot-interaction (HRI).
linkedin.com/posts/shalini-d…
shengze wang retweeted
Our AMRI team at NVIDIA Research is looking for Research Interns for 2027!
We’re building Interactive Embodied AI for dynamic 4D worlds, with research spanning AI, robotics, world models, neural rendering & human behavior understanding.
Interested? Email us with the subject “Seeking Research Internship 2027”.
Details: research.nvidia.com/labs/amr…
shengze wang retweeted
NVIDIA Research is looking for 2027 PhD interns! If you are interested in 4D reconstruction and/or world models, check out research.nvidia.com/labs/amr… and send me an email/DM!
Beautiful!
What if you could turn any number of photos (3, 8, 15, or even 60) into one clean 3D surface (pts & mesh) with Flow Matching?
Check out our new work, Surflo: Consistent 3D Surface Flow Model with Global State. 🧵
1/n
🔗anttwo.github.io/surflo/
2
77
📣 Interactive Physical AI (IPA) workshop @ CVPR
research.nvidia.com/labs/amr…
⏰ New Important dates:
Paper deadline: March. 7 on OpenReview
Notification: March 20
Camera-Ready: April 10
🎉 Announcing the first Interactive Physical AI Workshop at #CVPR2026.
Join us for a half-day workshop exploring AI systems that see, communicate, and act safely in our shared physical world — including robots, environment-aware avatars (e.g., AR telepresence), and on-device multimodal agents.
✅ Cross-disciplinary topics spanning vision, robotics, and multimodal AI
✅ Featuring invited speakers (incl. Yaser Sheikh @subail), poster sessions, and spotlight talks
📅 Paper deadline is Feb 28: openreview.net/group?id=thec…
More info: research.nvidia.com/labs/amr…
🙌 Organized by: @swookpark, @amritamaz, @mct1224, @lmathur_, @luminohope and @shalinidemello
💡 Sponsored by NVIDIA.
We look forward to seeing you at CVPR.
314
shengze wang retweeted
Can humanoids perform agile, autonomous, long-horizon parkour—based on what they see in the world?
We present 𝗣𝗲𝗿𝗰𝗲𝗽𝘁𝗶𝘃𝗲 𝗛𝘂𝗺𝗮𝗻𝗼𝗶𝗱 𝗣𝗮𝗿𝗸𝗼𝘂𝗿 (𝗣𝗛𝗣): a framework that chains dynamic human skills using onboard depth perception for long-horizon traversal.
1/6
shengze wang retweeted
🚀 Excited to share AnchorWeave — a local-memory-augmented framework for world-consistent long-horizon video generation.
- Global 3D reconstruction as memory accumulates cross-view misalignment and contaminates conditioning signals.
- We replace a single noisy global 3D memory with multiple retrieved local 3D memories and learn to weave them.
- Stronger long-horizon scene consistency and generalization ability.
🧵👇
📢 Interactive Physical AI (IPA) workshop @CVPR will invite some brilliant minds to discuss socially-intelligent #robots, environment-aware #avatars (e.g., AR telepresence), and on-device multimodal #agents.
Bookmark the workshop!
research.nvidia.com/labs/amr…
Paper deadline is 2/28
📢IPA 2026 (Interactive Physical AI) @CVPR 2026.
Announcing the 1st Interactive Physical AI workshop at CVPR 2026! This half-day workshop will bring together researchers to discuss AI systems that perceive humans and scenes, communicate via multimodal signals, and act safely in our shared physical world. 🧵
#CVPR2026
1
275
Check out our team's SIGGRAPH Asia Emerging Technologies demo coming to HongKong Dec 16-18! @SIGGRAPHAsia
excited to announce ▶️Play4D: efficient dynamic 4D neural scene reconstruction, for physical AI, VR, and light field displays!
see it live at #SIGGRAPHAsia2025 Emerging Technologies in Hong Kong!
🔗 research.nvidia.com/labs/amr…
shengze wang retweeted
Gemini Nano Banana Pro can solve exam questions *in* the exam page image. With doodles, diagrams, all that.
ChatGPT thinks these solutions are all correct except Se_2P_2 should be "diselenium diphosphide" and a spelling mistake (should be "thiocyanic acid" not "thoicyanic")
:O
shengze wang retweeted
This paper from Tsinghua University and Shanghai Jiao Tong University received perfect scores (6, 6, 6, 6) at NeurIPS 2025!
It aims to answer a key question: Does reinforcement learning really make large language models better reasoners?
The authors study Reinforcement Learning with Verifiable Rewards (RLVR) and find that while it improves accuracy for small k, it doesn’t create new reasoning patterns—meaning the base model still determines the upper limit of reasoning ability.
Across six RLVR variants, performance gains plateau, suggesting that current RL setups mainly refine reasoning rather than reinvent it.
Interestingly, it’s distillation, not RL, that shows genuine signs of emergent reasoning.
This research points to the next frontier for truly self-improving large language models.
Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?
Paper: arxiv.org/abs/2504.13837
Page: limit-of-RLVR.github.io
Our report: mp.weixin.qq.com/s/2-GDxs8j1…
📬 #PapersAccepted by Jiqizhixin
Thrilled that our paper received the only perfect score at NeurIPS this year.
Huge thanks to my collaborators and the reviewers. See you in San Diego!
limit-of-rlvr.github.io
papercopilot.com/statistics/…
credit to @papercopilot
shengze wang retweeted
Glad introduce our #ICCV2025 work Ponimator: Unfolding Interactive Pose for Versatile Human-human Interaction Animation
Website: stevenlsw.github.io/ponimato…
Paper: arxiv.org/abs/2510.14976
Code: stevenlsw.github.io/ponimato…
Poster: Wed 10.22, 10:30am #1109
@chuan_guo92603 @_JianWang_

























