Wenhao Zhang

EFlow's innovative separation of evidence retrieval and reasoning processes leads to substantial improvements in long-video reasoning performance, addressing critical biases in existing frameworks.

DrivingDepth achieves state-of-the-art depth estimation by leveraging sparse LiDAR to fine-tune pixel-wise scale without sacrificing geometric coherence.

Generative retrieval gets a reasoning upgrade: ThinkGR interleaves chain-of-thought with docid generation, boosting multi-hop retrieval performance by 6.86%.