OpenWAM: An Open, Modular Exploration Towards Systematic World-Action Model Pretraining Paper • 2609.07398 • Published Sep 7 • 78
LLaDA-Image: Building Strong Image Generators with Fully Open Training Recipes Paper • 2609.03796 • Published Sep 3 • 181
ZimaBlue: Evolving Generalizable World Action Models through Scalable Video Pre-training Paper • 2609.00188 • Published Aug 31 • 53