๐ค๐๐ | Senior undergrad, Yao class @Tsinghua_Uni | RA @uwcse | RS @wuji_global | Prev. RA @mldcmu.
Seattle, WA
- Tweets 326
- Following 1,577
- Followers 1,643
- Likes 1,108
Pinned Tweet
๐ค World Action Models predict the future to act better. However, WAM research remains fragmented in code, ad hoc in design, and unprincipled at scale.
๐ก Today, we release ๐ข๐ฝ๐ฒ๐ป๐ช๐๐ : ๐๐ป ๐ข๐ฝ๐ฒ๐ป, ๐ ๐ผ๐ฑ๐๐น๐ฎ๐ฟ ๐๐
๐ฝ๐น๐ผ๐ฟ๐ฎ๐๐ถ๐ผ๐ป ๐ง๐ผ๐๐ฎ๐ฟ๐ฑ๐ ๐ฆ๐๐๐๐ฒ๐บ๐ฎ๐๐ถ๐ฐ ๐ช๐ผ๐ฟ๐น๐ฑโ๐๐ฐ๐๐ถ๐ผ๐ป ๐ ๐ผ๐ฑ๐ฒ๐น ๐ฃ๐ฟ๐ฒ๐๐ฟ๐ฎ๐ถ๐ป๐ถ๐ป๐ด to change this.
๐ openwam-official.github.io/
๐งต1/14
6
40
263
2,264,756
๐ค World Action Models predict the future to act better. However, WAM research remains fragmented in code, ad hoc in design, and unprincipled at scale.
๐ก Today, we release ๐ข๐ฝ๐ฒ๐ป๐ช๐๐ : ๐๐ป ๐ข๐ฝ๐ฒ๐ป, ๐ ๐ผ๐ฑ๐๐น๐ฎ๐ฟ ๐๐
๐ฝ๐น๐ผ๐ฟ๐ฎ๐๐ถ๐ผ๐ป ๐ง๐ผ๐๐ฎ๐ฟ๐ฑ๐ ๐ฆ๐๐๐๐ฒ๐บ๐ฎ๐๐ถ๐ฐ ๐ช๐ผ๐ฟ๐น๐ฑโ๐๐ฐ๐๐ถ๐ผ๐ป ๐ ๐ผ๐ฑ๐ฒ๐น ๐ฃ๐ฟ๐ฒ๐๐ฟ๐ฎ๐ถ๐ป๐ถ๐ป๐ด to change this.
๐ openwam-official.github.io/
๐งต1/14
6
40
263
2,264,756
๐ฆ We release OpenWAM as a full stack empowering WAM research:
- Modular infrastructure for training, inference, deploy, and eval.
- OpenWAM-ฮฑ pretrained and post-trained weights.
- Study checkpoints for exploring the design choices.
- Evaluation protocols and data recipes.
๐ฅ We hope researchers can use the stack to reproduce the results, adapt OpenWAM-ฮฑ to new tasks and embodiments, and test new ideas for representations, worldโaction interaction, and embodied pretraining.
๐ญ There is much more to explore. We hope OpenWAM-ฮฑ serves as a strong, reproducible baseline and OpenWAM-Infra makes that next round of experiments easier to build.
๐ Project: openwam-official.github.io/
๐ Paper: arxiv.org/abs/2609.07398
๐ป Code: github.com/OpenWAM-Official/โฆ
๐ค Models & data: huggingface.co/OpenWAM
๐งต13/14
1
1
276
๐ OpenWAM is a large undertaking, made possible by the dedication and collective effort of the entire team. Thank you to everyone who contributed to this project, couldn't have done it without y'all!
๐ค Iโm grateful for the opportunity to have co-lead this project with @YuranWang_rise. Special thanks to the amazing adviors: @zhaohang0124, @linshaonju, @HaoDong123, and @LotusSapphire for their advice and guidance throughout the project.
๐งต14/14
2
256
I still donโt get why people think code as policy is so compelling. Language models and their harness can be very useful for robots, but writing programs that directly output for example, eef actions, seems to me not the best interface.
5
2
46
4,077
Hey, maybe you are looking for our STARFlow-V (starflow-v.github.io/starfloโฆ) ๐ค๐ฒ and it is not diffusion.
Most โautoregressive videoโ is not autoregressive.
Nobody pretrains BERT, distills a GPT-like student, and calls that native. Video does the equivalent every day: train a bidirectional short-clip DiT, then add a causal mask and a few-step sampler to produce a faster version.
I wrote a short blog on what native would actually require:
waynejin0918.github.io/how-fโฆ
( Thanks to the TML blog for providing stylistic references, and to grok4.6 for helping me polish my content. )
2
7
67
10,235
"Weโre using language as a crutch to help the deficiencies of our vision systems to learn good representations from images and video." Yann Lecun, Lex Fridman Podcast #416.
Sorry, this thread is very cool, but I still believe in the long run, building models with intrinsic understandings of the world requires a more vision-centric approach than modern LLMs.
2
2
72
6,751
This one is impressive. "Behavior Prompting", or in-context learning of robotics from seeing demonstrations has a long history, and it's a good time to go down this memory lane. (1/n)
Introducing GEN-1.5, a one-shot learner.
It can learn new tasks in a few seconds. Show it what to do, and it generalizes.
This capability emerged from pretraining on physical data at scale, as a step towards our mission of building general intelligence for the physical world.
2
1
43
3,624
This year is a especially noise year for this field. There are multiple research more or less related to this coming out everywhere:
- Behavior Prompting Policy: behavior-prompting.github.ioโฆ
- HumanEgo: humanego-ai.github.io/
- RoboTTT: Context Scaling for Robot Policies: arxiv.org/abs/2607.15275
- WAM-TTT: Steering World-Action Models by Watching Human Play at Test Time: arxiv.org/abs/2607.06988
5/n
1
3
248
I would end by a couple closing remark:
1. In-context learning for robotics is a long-standing problem;
2. Scaling context length (& format diversity) is a commonly-suggested advise for robotics. However, does/should this icl/reasoning capability happen at the VLA layer rather than the VLM layer? Since VLA layer inference needs to be served at a certain frequency for closed-loop control, and naive attention's compute scale quadratically with context length;
3. If you combine this half-a-year+ training with DYNA-2's curve, the takeaway should not be let's scale to 1B hours of robot data & 1 million gpus for compute, but there's some fundamentally wrong with the current paradigm, and there's something crucial missing in the robot learning repertoire.
What an exciting time to be alive.
6/6
6
678