Compressed Convolutional Attention: Efficient Attention in a Compressed Latent Space Paper • 2510.04476 • Published Oct 6, 2025 • 17
LatentLoop: Scaling Looped MoE Models with Asymmetric Compute Allocation Paper • 2610.19975 • Published 1 day ago • 56
LeWAM: End-to-End World and Action Modeling with JEPAs Paper • 2610.12818 • Published 8 days ago • 65
Urbex: Agentic Spatial Grounding for City-Scale 3D Scenes Paper • 2610.18356 • Published 3 days ago • 114
RSI-Anything: Unlocking Recursive Self-Improvement with World Models Paper • 2610.13912 • Published 7 days ago • 46