Shengyuan Ding

multimodal large language model (3) instruction following (2) direct preference optimization (2) supervised fine-tuning (2) preference alignment (1) vision language model (1) reward model (1) multi-modal large language model (1) proximal policy optimization (1) human preference alignment (1) test-time scaling (1) context-aware generation (1) creative generation (1) creative capabilities (1) reinforcement learning (1) multimodal reward (1) benchmark evaluation (1) visual question answering (1) preference learning (1)