Fan Feng (@ffeng01) profile picture

Postdoc @UCSanDiego, @mbzuai | working on world models | Previously @CityUHongKong | Visitor @UvA_Amsterdam

Joined November 2016
Photo by Fan Feng (@ffeng01)Photo by Fan Feng (@ffeng01)Photo by Fan Feng (@ffeng01)
Fan Feng retweeted
Doing a PhD and interested in World Models? Come work with me and the Meta FAIR World Model team for a year. Exciting research, hundreds of GPUs and free food are awaiting you.
5
8
1
296
13,736
Fan Feng retweeted
🤖 Understanding what exactly happens during your robot policy deployments is key to improving it. But interpreting raw rollouts is hard because they lack structure. Introducing Behavior Graphs: turn many rollouts into an interpretable graph of what the policy is doing. 🧵👇
2
21
2
87
7,886
Fan Feng retweeted
Can more pretraining improve a robot policy’s success rate, yet makes it worse for porttraining? 🤔 In our new work, we find that it can. Continued supervised pretraining of generative robot policies can improve reliability at the expense of coverage, and this tradeoff affects downstream posttraining. Read on to see why this happens and how we can preserve coverage. 1/n 🧵
17
57
9
363
36,500
What should a robot remember? Can it learn to remember better than we tell it to? Excited to share MEMI: Reverse-Engineering What Robots Should Remember 📈 97.0% end-to-end, 99.4% w/ low-level controller on RoboMME, nearly solving it for the first time 🧵👇
7
20
6
137
17,362
🚨 Stop feeding your world model the same boring demos. Switch to free-range play data! 🚨 Our CoRL 2026 paper finds we can optimize autonomous data collection to focus on missing action consequences, achieving up to 85x the data scaling efficiency of expert data collection!🧵
8
33
9
154
16,189
If video models scale like LLMs, they'll soon generate a perfect video plan for any embodied task. Do we need anything else? Introducing Mirror, which learns to follow these plans just by practicing, and even crafts an iron pickaxe in Minecraft w/o rewards or action labels! (1/n)
9
44
14
393
54,217
Why learn everything from scratch when centuries of science have already taught us how the world works? We’re excited to introduce ReRoute 🔀: bringing what-if reasoning to scientific AI models by combining observational data with partial mechanistic knowledge. 🔥 No controlled intervention data required. 🧵 1/7
Excited to share ReRoute, our new framework for scientific counterfactual prediction: informed by partial physics knowledge for much better generalization. 🌍 AI models can be remarkably accurate on familiar data, yet struggle with scientific ‘what-if’ questions The key idea: Combine observational data with partial physical knowledge to predict unseen interventions—without requiring new controlled experiments. Instead of retraining models on expensive new simulations, ReRoute uses known partial physical mechanisms to guide pretrained models through lightweight fine-tuning on existing data. The results: 🌍 Up to 31.8% lower prediction error under severe CO₂ distribution shifts in climate emulators. 🔬 Up to 89.6% lower counterfactual error in controlled advection–diffusion experiments. ⚡ No new intervention data. No retraining from scratch. Just lightweight fine-tuning. We also establish a causal identification result under explicit structural assumptions, with a machine-checked proof in Lean. The bigger picture: Scientific AI doesn't always need more data or bigger models to generalize beyond what it has seen. Sometimes, knowing just a little about how the world works can make a big difference. @Dingling_Yao @kahaan_gandhi @FrancescoLocat8 @Caltech 📄 Paper: arxiv.org/abs/2610.02252 #AIforScience #ScientificML #ClimateAI #CausalInference #PhysicsAI
1
13
33
2,305
Fan Feng retweeted
Robotics has spent years scaling demonstration data and model size. What happens when we scale EXPLORATION for robot agents? We're incredibly excited to share RAX, an agentic robotics system that learns by scaling online exploration. A coding agent connects directly to the robot’s perception and control APIs, learns from feedback, and builds programs that perceive, plan, and act. The model stays frozen. As exploration scales, the agent improves its programs and builds up reusable skills. The results: • ZERO teleoperation or action demonstrations on real robots—from insertion with 0.1 mm clearance and air-fryer manipulation to deformable objects and long-horizon mobile whole-body control. • +37 percentage points on average over leading learned policies across five simulation benchmarks: RoboDojo, BEHAVIOR-1K, HumanoidArena, DexJoCo’s bimanual tasks, and LIBERO-Pro. • One framework spanning single-arm, bimanual, dexterous, humanoid, and mobile robots, with new embodiments connected through APIs. The craziest part: on each real-world embodiment, going from scratch to mastering the tasks took roughly two hours. No teleop. No model finetuning. We’re open-sourcing the full framework AND every experiment session. We can’t wait to see what people build with it! Project + sessions: robot-agent-exploration.com/ Paper: assets.robot-agent-explorati… GitHub: N Project fully led by @RunhanH and @xczhangxx who are absolutely awesome. How it works, experiments, and robot videos below 🧵 (1/N)
41
104
32
589
77,470
Want to know how well agents can really control robots directly? In RoboARCade, we systematically evaluate and diagnose frontier agents controlling robots directly using tens of thousands of dollars’ worth of agent trajectories. RoboARCade contains diverse, precise, and long-horizon tasks. Paper, code, assets and every trace are out today. roboarcade.org
10
42
9
172
17,682
Fan Feng retweeted
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/
27
84
12
731
69,128
Fan Feng retweeted
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/
13
48
9
359
40,491
Fan Feng retweeted
TC-WM is accepted at #NeurIPS2026! 🎉 A generic visual representation isn’t always a good latent state for planning and control. For control, we need more than visual fidelity. Instead, we aim for task-centric representations that reflect physical dynamics and task-relevant information, while still leveraging the scalability of visual foundation models. Compared with DINO-WM, TC-WM learns a lower-dimensional representation with higher latent-space utilization and diversity. These improvements translate into stronger performance, especially with the clearest gains on 3D contact-rich manipulations. Webpage: minghaofu.com/tc-wm/ Paper: arxiv.org/abs/2605.25620 Code: github.com/MinghaoFu/TC-WM
4
11
1
139
9,768
I am honored to have received runner up best paper award at ICML's forecasting-workshop.github.… for my Bayesian forecasting paper. In my talk I showed some unpublished results on a more rigorous Bayesian approach. I have now added these results to the arxiv paper....
9
21
1
363
27,248
The strongest healthcare LLM, custom-built for your enterprise, owned by you🌸 Meet @actAVAai Cura: 1T agentic model trained by recursive self-improvement for long clinical + health admin workflows Try: actava.ai/cura Share your use case: $20 credits + early access👇
128
172
196
691
1,145,851
Fan Feng retweeted
Heading to ICML 2026 next week! I'll be presenting our work: On Information Self-Locking in Reinforcement Learning for Active Reasoning of LLM Agents 📍 Poster July 8 (Wed), 2:30–4:15 PM KST Hall A #2014 The paper studies a failure mode of outcome-based RL for active reasoning, information self-locking, and introduces AREW, a simple advantage reweighting approach to improve information-seeking and belief-updating behaviors. If you're attending ICML, feel free to stop by for a chat. I'd be happy to discuss active reasoning, RL, LLM agents, or related ideas. See you in Seoul!
5
29
1,925
Nice work on self-consistent world models! Just one small note on the positioning: a couple of papers have studied ideas around forward–inverse consistency in world models, e.g. our recent WAV and the earlier RLIR from MSR
New work with @nvidia: evaluating robot policies entirely inside a world model. The policy acts, the model imagines the consequences, and the imagined evals predict real-world results. 🧵 real vs world-model rollout side by side📷
2
12
48
8,514
Vision foundation models predict hurricane intensity from satellite imagery remarkably well — and pass out-of-distribution tests. Yet their internal representations collapse exactly where physics matters most. We call this the Perception–Physics Paradox. New at ICML 2026 🧵
3
13
1
44
6,772
Introducing CHI-Bench on @huggingface: the world’s first long-horizon healthcare benchmark for AI agents. 75 real healthcare workflows + 20 apps + 200+ MCP tools + 1,290 skills + process / outcome rewards huggingface.co/datasets/acta… Any questions, lmk!
8
20
7
142
32,886
Under review? Still welcome! The RLBrew deadline is 1 week away — great place to get early feedback on your work!
Reminder! RLBRew deadline in coming up in 7 days! Submit your works soon👩‍💻 Reminder that we accept under review papers! This is a good place to discuss your ideas and get feedback from the community
1
5
6
3,368
(1/n) After a few months of work with multiple hospitals, universities and research facilities, today we're open-sourcing CHI-Bench: the first long-horizon benchmark for healthcare AI agents on real clinical and healthcare workflows. Best frontier agent overall: 28% pass@1. End-to-end prior authorization: 0%. A thread on what we found 🧵
10
7
2
18
782