Nan Huang

A significant reasoning-generation gap exposes the limitations of current generative models, revealing that they often produce outputs that are locally plausible yet globally illogical.

LLMs may excel at predicting outcomes but often fail to grasp the underlying scientific validity of the laws they propose, revealing critical gaps in their discovery capabilities.

Unlock 36% better video depth estimation and 20% better camera pose estimation by simply letting your model learn from its own unlabeled video predictions.