4 papers indexed
Utkarsh A. Mishra, Yongxin Chen, Danfei Xu, Yang Liu, Xi Chen, Jiayuan Mao
Generative video foundation models exhibit strong compositional priors, yet world-action models (WAMs) and video-action models (VAMs) often lose these priors after finetuning on robotic action data. We refer to this discrepancy as the video-action generalization gap. In this pape…
Zelin Zhao, Min Shi, Bo Yuan, Haotian Xue, Jialuo Li, Lama Moukheiber, et al.
World models aim to capture environment dynamics in ways that support perception, reasoning, and action, and have recently become a central direction in Vision-Language-Action-World (VLAW) modeling. Meanwhile, unified vision-language models have demonstrated strong multimodal gen…
Zinuo Chang, Yipu Chen, Byoungwoo Park, Hongzhe Yu, Yongxin Chen
Diffusion-based motion planners have demonstrated strong performance in generating diverse and high-quality robot trajectories in cluttered environments with multiple feasible solutions. However, existing approaches typically operate on fixed-length waypoint sequences, making the…
Embedded finance is usually described as the integration of financial services into non-financial platforms. This article argues that the deeper regulatory transformation is the embedding of future legal and economic obligations into ordinary consumption, work, business, and ente…