Cola2: Co-Adaptive Hierarchical Language Modeling with Continuous Semantic Compression and Balanced Discrete Generation Paper • 2610.31305 • Published 1 day ago • 14
Frechet Forcing: Training Long-Horizon Streaming World Models from Short Clips Paper • 2610.25666 • Published 5 days ago • 22
MiniWorld: Democratizing the Training of Video World Models from Scratch Paper • 2608.01127 • Published Aug 2 • 20