Author

Leyi Wu

Observed name: Wu, Leyi

Institutions: Hong Kong University of Science and Technology, Kuaishou (China), The Hong Kong University of Science and Technology (Guangzhou)

Observed institutions: Hong Kong University of Science and Technology, Kuaishou (China), The Hong Kong University of Science and Technology (Guangzhou)

OpenAlex ID: A5123915201

A Mechanistic View on Video Generation as World Models: State and Dynamics

2026 · Wang, Luozhou, Chen, Zhifei, et al. · arXiv (Cornell University) · 10

Motion Forcing: A Decoupled Framework for Robust Video Generation in Motion Dynamics

2026 · Xu, Tianshuo, Chen, Zhifei, et al. · arXiv (Cornell University) · 5

STANCE: Controllable Video Generation for Structured Dynamics via Sparse-To-Dense ANChored Encoding

2026 · Zhifei Chen, Tianshuo Xu, et al. · Lecture notes in computer science · 5

RoboStressBench: Benchmarking VLM Robustness to Physical Visual Stress in Embodied Scenes

2026 · Wu, Leyi, Zhao, Yifan, et al. · arXiv (Cornell University) · 3

In-Context Learning for Robots: Methods and Applications

2026 · Huang, Haojian, Li, Zexi, et al. · arXiv (Cornell University) · 2

AnimationBench: Are Video Models Good at Character-Centric Animation?

2026 · Wu, Leyi, Fang, Pengjun, et al. · arXiv (Cornell University) · 2

RoboAware: Learning to Coordinate Embodied Skills from Counterfactual Outcomes

2026 · Zhou, Bohan, Chen, Xingbei, et al. · arXiv (Cornell University) · 1

The Right Inference Strategy Is All You Need: Nearly Training-Free Domain-Wise Inference for EgoCross Challenge

2026 · Wu, Leyi, Zhao, Yifan, et al. · arXiv (Cornell University) · 1

Anvil: Scaling Embodied Spatial Intelligence through Knowledge Recirculation

2026 · Ruan, Weilin, Zhou, Bohan, et al. · arXiv (Cornell University)

Seg3DParts: Segmentation-Grounded Controllable Part-Level 3D Generation

2026 · Lin, Jiantao, Chen, Meixi, et al. · arXiv (Cornell University)

The First EgoCross Challenge at EgoVis 2026: Cross-Domain Egocentric Video Question Answering

2026 · Fu, Yuqian, Qian, Tianwen, et al. · arXiv (Cornell University)

Works count: 11

Citation count: 29

H-index: 3

I10-index: 1