FantasyVLN: Unified Multimodal Chain-of-Thought Reasoning for Vision-Language Navigation Paper • 2601.13976 • Published Jan 20 • 21