1 paper · 1 filter
Zefan Cai, Haoyi Qiu, Tianyi Ma +9
Video foundation models generate visually realistic and temporally coherent content, but their reliability as world simulators depends on whether they capture physical, logical, an…