Ph.D. Student @UCSanDiego | World Model | Representation Learning |
San Diego, California, USA
Joined June 2023
- Tweets111
- Following275
- Followers320
- Likes1.1K
Pinned Tweet
JEPAs can 𝐧𝐨𝐰 jointly imagine what happens and what to do next!
Meet 𝐋𝐞𝐖𝐀𝐌: A JEPA for the real world with
32.3× faster planning, 19 ms per plan
Improves success 28.6% → 89.7% on contact-rich manipulations
But also ... 🫶 a WAM for everyone with
17M · 1 GPU · 1 hyperparameter
Works on a real robot, not just a simulation !!
👉 le-wam.github.io/
Minghao Fu📌CoRL retweeted
hyped to be presenting LeWAM as a poster at the #CoRL2026 workshop do-robots-need-world-models.…!
check out our poster if you're interested in applying JEPA for robot learning
Heading to Austin next week for #CoRL!
Excited to present LeWAM at the CoRL workshop! do-robots-need-world-models.…
Always happy to grab a coffee and chat about robot learning and world models. Feel free to reach out! ☕
JEPAs can 𝐧𝐨𝐰 jointly imagine what happens and what to do next!
Meet 𝐋𝐞𝐖𝐀𝐌: A JEPA for the real world with
32.3× faster planning, 19 ms per plan
Improves success 28.6% → 89.7% on contact-rich manipulations
But also ... 🫶 a WAM for everyone with
17M · 1 GPU · 1 hyperparameter
Works on a real robot, not just a simulation !!
👉 le-wam.github.io/
Minghao Fu📌CoRL retweeted
Really excited to share Catch It All!, my first paper as last author! In this work led by @Kehlani_Fay @arth_shukla we develop a system for catching diverse objects using a single egocentric camera 🧵👇
Minghao Fu📌CoRL retweeted
Coding agents can build an app. What if they could give it a body?
Meet Metatron: a versatile System 1 controller for humanoids. A goal-based interface connects System 2 planners, including coding agents, to whole-body movement.
A few examples below. 1/5
Minghao Fu📌CoRL retweeted
World to World
I gave Claude Opus 5.5 a Gaussian splat of a garden and asked it to rebuild the scene in Blender
Procedural objects, Geometry Nodes, Shader Nodes, then captured textures mapped onto selected surfaces
True lol, can’t stop ppl from naming their work whatever they want. And tbh, a lot of ppl pumping out AI slop couldn’t care less about their reputation
in an interesting followup...
seems like the Latent evolving World Action Model paper from baidu, initially submitted on sep. 23, only added the "LeWAM" label to the paper title on oct. 10, literally 2 days after a couple of LeWAM papers got submitted (oct. 8), one of which is a followup to @randall_balestr's line of work tracing back to LeJEPA and LeWorldModel, and the other is parallel work from nvidia (which arguably popularized the WAM acronym) building on LeWM in a different direction.
6
1,206
Minghao Fu📌CoRL retweeted
Today, Dream Labs is coming out of stealth with Dream-1, a robot foundation model that controls bimanual robots it has never seen, on tasks it was never shown, in places it has never been, from $3/hour.
Dream API is live for everyone, today. 🧵
What exactly is the difference between a goal-conditioned policy and “search-free planning with a world model”? Am I missing some new terms for policies 😅?
It’s been 3+ months since INTACT dropped — and it’s already been cited 10+ times 🚀🔥
Huge congrats to HIJP 👏 Really glad to see people learning INTACT, using INTACT, and going beyond INTACT ✨
Real innovation grounded in JEPA theory is rare — big shoutout to the JEPAs 💪
INTACT is still one of the SOTA works on Direct 🎯📈
So… is it time for me to bring you all “some” surprises too? 👀Time for another real shockwave ⚡️
1
11
1,068
Minghao Fu📌CoRL retweeted
Representation conditioning is better than representation alignment.
The two recent papers reported similar observations. Time for some new trials?
RECON: arxiv.org/html/2610.20608v1
Dino forcing: arxiv.org/abs/2610.11751v1
Minghao Fu📌CoRL retweeted
DINOv2 has never seen a caption, and Qwen3 has never seen an image. We still aligned their embedding spaces without a single image-caption pair.
It even works when the images and the captions come from different datasets.
Project page: dominik-schnaus.github.io/un…
⬇️
Minghao Fu📌CoRL retweeted
If end-to-end is the question, SIGReg is the answer! Great work pushing end-to-end pretraining for WAMs! Also an important step for JEPAs to be able to plan and solve tasks at lightning speed!
JEPAs can 𝐧𝐨𝐰 jointly imagine what happens and what to do next!
Meet 𝐋𝐞𝐖𝐀𝐌: A JEPA for the real world with
32.3× faster planning, 19 ms per plan
Improves success 28.6% → 89.7% on contact-rich manipulations
But also ... 🫶 a WAM for everyone with
17M · 1 GPU · 1 hyperparameter
Works on a real robot, not just a simulation !!
👉 le-wam.github.io/
JEPA control on real robots: LeWAM executes action chunks directly without goal image, and is competitive with other policies and VLAs
1
9
1,259
Joint work with @tavis_siebert and @ErykHalickii, and huge thanks to @randall_balestr for the guidance and support! 🙏
🌐 Project: le-wam.github.io/
💻 Code: github.com/MinghaoFu/lewam
🤗 Data & Checkpoints: huggingface.co/LeWAM
I would say provably learning atomic latent variables is underappreciated in world models. Check out this paper for a fresh perspective!
Is it possible to provably recover individual latent variables of the true world, even without reconstruction (e.g., JEPA)?
Yes, with DSReg, a simple regularization that can be applied post hoc to your pretrained model!
dsreg.github.io/
7
1,659
Hmmm… the day has finally come
We’re releasing a broad range of new mathematical results produced by an internal frontier model.
We’ve been consulting with the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study, and we have drawn on their advice and public recommendations to inform how we release these results.
github.com/openai/math
1
397