Hao Li

Over 62% of AI supply chains include artifacts with no declared license, raising serious concerns about compliance and accountability in AI development.

Achieving 93.8% accuracy in predicting molecular structures from multimodal spectroscopic data could revolutionize how chemists approach organic synthesis and analysis.

IGGT4D achieves unprecedented consistency in geometry-instance representation from streaming video, outperforming existing methods in dynamic scene understanding.

A simple training framework boosts pixel-level tampering detection performance in VLMs by over 26%, showcasing the power of balanced sampling and late-injection strategies.

Achieve SOTA in long-horizon spatial understanding by training a model to streamingly maintain and update spatial evidence from video via test-time adaptation of fast weights.