Feng Yao

DenseMixer: Improving MoE Post-Training with Precise Router Gradient

Contamination Detection for VLMs Using Multi‑Modal Semantic Perturbations

TemporalBench: Evaluating Fine-Grained Temporal Dynamics Understanding for Multimodal Models

Finish First, Perfect Later: Test-Time Token-Level Cross-Validation for Diffusion Large Language Models

Training Language Models to Generate Quality Code with Program Analysis Feedback

TemporalBench: Towards Fine-grained Temporal Understanding for Multimodal Video Models