@wanggrun

Chief Scientist at X-Era AI. Associate Professor and Young Scientist at Sun Yat-sen University. Former Research Fellow at the University of Oxford.

Joined April 2012
  • Tweets37
  • Following132
  • Followers39
2Media
GPT meets control theory; how can GPT control a robot?
1
1
44
Guangrun Wang retweeted
Guangrun Wang@wanggrunJul 4
When Will Physical AGI Arrive? Here’s What Our Team Is Doing
1
1
76
When Will Physical AGI Arrive? Here’s What Our Team Is Doing
1
1
76
Guangrun Wang retweeted
Guangrun Wang@wanggrunFeb 23
Replying to @fly51fly
This paper has been accepted to CVPR 2026: Weiqi Li, Quande Zhang, Ruifeng Zhai, Liang Lin, Guangrun Wang, VLA Models Are More Generalizable Than You Think: Revisiting Physical and Spatial Modeling
1
1
1
113
Guangrun Wang retweeted
Robotics Papers@OWWFeb 12
RADAR: Benchmarking Vision-Language-Action Generalization via Real-World Dynamics, Spatial-Physical Intelligence, and Autonomous Evaluation Yuhao Chen, Zhihao Zhan, Xiaoxin Lin, Zijian Song, Hao Liu, Qinhan Lyu, Yubo Zu, Xiao Chen, … arxiv.org/abs/2602.10980 [𝚌𝚜.𝚁𝙾]
1
1
1
287
VLA Models Are More Generalizable Than You Think: Revisiting Physical and Spatial Modeling. arxiv.org/abs/2512.02902

VLA Models Are More Generalizable Than You Think: Revisiting...

Vision-language-action (VLA) models achieve strong in-distribution performance but degrade sharply under novel camera viewpoints and visual perturbations. We show that this brittleness primarily...

arxiv.org
2
47
To address this, we propose a one-shot adaptation framework that recalibrates visual representations through lightweight learnable updates. We apply a global affine transformation to visual tokens and improves Libero viewpoint accuracy from 48.5% to 87.1% with only 4K parameters.
32
Vision-language-action (VLA) models achieve strong in-distribution performance but degrade sharply under novel camera viewpoints and visual perturbations. We show that this brittleness primarily arises from misalignment in Spatial Modeling, rather than Physical Modeling.
42
arxiv.org/abs/2510.01047 Authentic Discrete Diffusion Model

In-Situ Tweedie Discrete Diffusion Models

While diffusion models excel at generating continuous data such as images, adapting them to discrete tasks has relied on indirect approaches that either operate in continuous embedding spaces or...

arxiv.org
2
1
1
98
more replies
2.Introducing time-weighted cross-entropy loss during training to avoid the over-smoothing problem caused by MSE; 3. Using an argmax + one-hot feedback iteration during inference to achieve true step-by-step discrete denoising.
1
80
For the first time, this mechanism faithfully transplants the principles of diffusion into the world of discrete symbols, achieving true “symbol-level diffusion.”
45
Are diffusion language models all wrong? SYSU team proposes the first authentic discrete diffusion model
75
Guangrun Wang retweeted
Guangrun Wang@wanggrun2 Oct 2025
arxiv.org/abs/2510.01047 Authentic Discrete Diffusion Model

In-Situ Tweedie Discrete Diffusion Models

While diffusion models excel at generating continuous data such as images, adapting them to discrete tasks has relied on indirect approaches that either operate in continuous embedding spaces or...

arxiv.org
2
1
1
98
Load more