Zhe Chen

Bridge the gap between offline model scaling and online deployment in recommendation systems: Rec-Distill enables lightweight student models to capture a substantial portion of the performance gains from massive teacher models.

Omni-LLMs struggle to identify the same objects across different modalities, but a new dataset and training strategies can significantly improve their cross-modal reasoning.

Current methods to protect satellites from radiation drain batteries and interrupt service, but a new routing protocol can minimize both.