
Visiting PhD student at @StanfordSVL. Working with Prof. @drfeifei. PhD student of NUS. Prev. Tsinghua Univ. Roboticist.
Stanford
Joined August 2021
- Tweets579
- Following1K
- Followers910
- Likes763
Pinned Tweet
Tired of pick-and-place benchmarks?
Want to explore manipulation tasks of different difficulty levels?
Introducing ManiLadder 🪜, a simulation manipulation benchmark that organizes 112 simulated tasks into 4 calibrated difficulty levels × 8 suites. 🧵
maniladder-benchmark.github.…
Does a ladder calibrated in simulation hold on real robots?
We rebuilt 8 tasks in the real world. Success by level:
90% → 40% → 10% → 0%
The difficulty order carries over.
1
1
83
All open: 112 tasks, demos, all checkpoints.
You can use it to study any topic you want! Continual learning, transfer learning, cross-embodiment, ...
Paper: arxiv.org/abs/2611.16374
Website: maniladder-benchmark.github.…
Code: github.com/HeegerGao/ManiLad…
2
85
DreamCan (dreamcan.github.io/) won the best paper finalist of CoRL workshop on Compositional and Modular Learning in the Era of Scaling in Robotics!
🤖Robots need world models for visual planning, but they don't need to imagine every frame.
Introducing DreamCan: Sparse Visual Planning for Long-Horizon Robot Manipulation.
Project: dreamcan.github.io/
Code: github.com/HeegerGao/DreamCa…
Paper: arxiv.org/abs/2610.24981
So proud of what the community achieved this year! It's been so much fun building the challenge and seeing everyone's creative solutions. Huge congrats to the winners, and thank you to every team who took part! Come say hi at CoRL this week 👇
The results for the 2nd BEHAVIOR Challenge at CoRL 2026 are out! 🚀 In just one year, here's how far we've come together:
🏠 2× benchmark scale: 50 → 100 tasks, 3 → 7 scenes
👥 3× participation: 96 → 265 participants, 18 → 57 submissions
📈 4× full-task success: 12% → 48%
Congratulations to our winners!
🥇 The North Star, Huawei Noah EAI Team
🥈 IF1001, Intellifusion
🥉 Cyber, Xiaomi Robotics
Thank you to every team, from academia, industry, and independent labs worldwide, for making this year's challenge exceptional! 🙌
📝 Challenge description: teapaw.org/drfeifei/status/207672…
(1/N)
Chongkai Gao retweeted
Very excited to show this demo at CoRL!
Policy? World model? Value learner (temporal distance)? Why not treat them all as prompts, trained with just next token prediction?
Turns out it's a surprisingly strong pretraining objective! We trained one 8B model in-house from scratch (not VLA or WAM), built from an amazing series of works from friends at @NeuroAILab, led by @AtlasKazemian @KlemenKotar
And it works zero-shot with new objects and goal images! Come try it out in JW Marriott Lone Star (3rd floor)!
Robot Capabilities as Probabilistic Inference over Next Token Prediction
We are doing a demo at CoRL debuting PsiBorg: an 8B universally promptable robot model for open-world manipulation!
One model works simultaneously as a policy, a simulator, and a progress tracker. It is trained from scratch, developed fully in-house at #Stanford, not a VLA or WAM, and (tries to) work with new objects and new goal images you throw at it!
Come try it at #CoRL2026!
🕛 Mon Nov 9 · 12–6:30 PM CST
🕙 Tue Nov 10 · 10 AM–6:30 PM CST
📍 Demo #2 · Lone Star (3rd floor), JW Marriott Austin 🌐 neuroailab.github.io/psi-web…
Very pround to be part of this demo!
Come try it at #CoRL2026!
🕛 Mon Nov 9 · 12–6:30 PM CST
🕙 Tue Nov 10 · 10 AM–6:30 PM CST
📍 Demo #2 · Lone Star (3rd floor), JW Marriott Austin
Robot Capabilities as Probabilistic Inference over Next Token Prediction
We are doing a demo at CoRL debuting PsiBorg: an 8B universally promptable robot model for open-world manipulation!
One model works simultaneously as a policy, a simulator, and a progress tracker. It is trained from scratch, developed fully in-house at #Stanford, not a VLA or WAM, and (tries to) work with new objects and new goal images you throw at it!
Come try it at #CoRL2026!
🕛 Mon Nov 9 · 12–6:30 PM CST
🕙 Tue Nov 10 · 10 AM–6:30 PM CST
📍 Demo #2 · Lone Star (3rd floor), JW Marriott Austin 🌐 neuroailab.github.io/psi-web…
Chongkai Gao retweeted
Just arrived in Austin for #CoRL2026! 🤠
Excited to present our work “Scale Up Strategically” this Tuesday! Would love to grab a chat!
📍 JW Marriott Austin | Poster #101
🗓️ Tuesday, Nov 10 | 5:00–6:30 PM
📌 Poster Session II
How should we design training data distribution to improve robot foundation models to generalize to open-world language instruction?
🤔Instead of simply scaling up data volume, we propose a smart data collection strategy to allocate demonstrations across the language instruction space.
🔍 Our data collection strategy has proved effective across more than 6 common robot foundation models🔥 and over 15 simulation and real-robot settings 🤖.
🔗 Website: yqi19.github.io/scale-up-str…
💻 Code: github.com/yqi19/CoRL2026-Sc…
🤗 Data: huggingface.co/datasets/yqi1…
Collaborated with #NVIDIA, 💗 thanks to my amazing collaborator: @HaojieHuang13 @boce_hu @XinyiXu0214 @zhang_yeee
#CoRL2026 #Robotics #Data
Heading to #CoRL2026! 🤖 Bringing DreamCan, ManiLadder & SteerAct, plus Demo 2 and a TAG/G2A oral.
Let’s chat world models, simulation benchmarks, agentic learning & hierarchical robotics. See you in Austin! 👋 – at Millbrae, CA
Chongkai Gao retweeted
Planners operate on task timescales. Feasibility changes on control timescales.
Bridging that gap takes a high-frequency mid-level layer. Introducing EPO: Energy-Based Policy Orchestration #CoRL2026
EPO gives each low-level policy an energy: how feasible a goal is, and how well the policy will reach it. Energies are differentiable and comparable across policies, so EPO optimizes the policy–goal pair at control rate.
Unlike a learned MoE router, EPO is compositional: adding a skill means training one new energy, nothing else. It's also more precise and switches policies 7× less often, pushing the precision–robustness frontier.
📄 Paper: arxiv.org/abs/2611.02546
🌐 Project: yewr.github.io/epo
💻 Code: github.com/YeWR/EPO
@wangyenjen @elvis_hsieh77 @frankzydou @ShiboZhaoSLAM @pabbeel @du_yilun @MIT_LISLab
Alright, alright, alright…
CoRL registration is completely maxed out, with hundreds of people on the waiting list. Please stop flooding my inbox for tickets 😂
You can still join the fun! Our partners are hosting social events all over town. We gathered the list of public events here:
corl.org/program/socials
Many don’t require a conference badge to attend. Let’s turn Ausitn into a Robotics Party Town!
CoRL is now sold out. We have an incredible program lined up. Can't wait to see y'all in Austin next week!
corl.org/program
















