Advancing Multimodal Reasoning: From Optimized Cold Start to Staged Reinforcement Learning Paper • 2506.04207 • Published Jun 4, 2025 • 48
