Yueyi Sun

Latest

• Mixed-r1: Unified reward perspective for reasoning capability in multimodal large language models
• Sa2va: Marrying sam2 with llava for dense grounded understanding of images and videos