MM-Zero: Self-Evolving Multi-Model Vision Language Models From Zero Data Paper • 2603.09206 • Published Mar 10 • 53
Omni-Diffusion: Unified Multimodal Understanding and Generation with Masked Discrete Diffusion Paper • 2603.06577 • Published Mar 6 • 49
Thinking to Recall: How Reasoning Unlocks Parametric Knowledge in LLMs Paper • 2603.09906 • Published Mar 10 • 75