FlowTok: Flowing Seamlessly Across Text and Image Tokens Paper • 2503.10772 • Published 21 days ago • 18
Beyond Next-Token: Next-X Prediction for Autoregressive Visual Generation Paper • 2502.20388 • Published Feb 27 • 15
COCONut-PanCap: Joint Panoptic Segmentation and Grounded Captions for Fine-Grained Understanding and Generation Paper • 2502.02589 • Published Feb 4 • 10
Democratizing Text-to-Image Masked Generative Models with Compact Text-Aware One-Dimensional Tokens Paper • 2501.07730 • Published Jan 13 • 17