Pengxiang Li

pengxiang

·

pixeli99

AI & ML interests

Video generation, Image editing, AD

Recent Activity

upvoted a paper 3 days ago

ZimaBlue: Evolving Generalizable World Action Models through Scalable Video Pre-training

upvoted a paper 30 days ago

GST-Bench: Can VLMs Develop Global Spatial Awareness from Video?

upvoted a paper about 1 month ago

Towards Physics of Multimodal Pretraining: Knowledge Flow, Modality Synergy, Early Unification, and Recipes

View all activity

Organizations

pengxiang 's papers 17

arxiv:2605.04647

arxiv:2602.23225

arxiv:2509.20109

arxiv:2508.19982

arxiv:2508.05731

arxiv:2507.20879

arxiv:2506.22049

arxiv:2506.21285

arxiv:2505.20199

arxiv:2504.14239

arxiv:2502.05795

arxiv:2501.04575

arxiv:2412.13795

arxiv:2411.17223

arxiv:2405.18380

arxiv:2404.10595

arxiv:2312.00651