AlignVLM: Bridging Vision and Language Latent Spaces for Multimodal Understanding Paper • 2502.01341 • Published Feb 3 • 36