3 ms·Towards General World Models for Video Generation via Predicting Masked Tokens1 points by famouswaffles 3y ago