vlmmultimodal [2410.04751] Intriguing Properties of Large Language and Vision Models LLaVA Molmo Pixtral Qwen2 VL Aria [2410.05993] Aria: An Open Multimodal Native Mixture-of-Experts Model Nvidia Eagle GitHub - NVlabs/EAGLE: EAGLE: Exploring The Design Space for Multimodal LLMs with Mixture of Encoders