Fangxun Shu

3

Papers

34

Total Citations

Papers (3)

LLaVA-MoD: Making LLaVA Tiny via MoE-Knowledge Distillation

MARS: Mixture of Auto-Regressive Models for Fine-grained Text-to-image Synthesis

Detecting and Mitigating Hallucination in Large Vision Language Models via Fine-Grained AI Feedback