Render to Reason: Novel-View Semantic Prediction Improves Spatial Understanding in VLMs Paper • 2610.05417 • Published 6 days ago • 1
ReToken: One Token to Improve Vision-Language Models for Visual Retrieval Paper • 2607.28627 • Published Jul 30 • 9
4Director: Controlling Video World Models with Rigid 3D Geometry Paper • 2610.02160 • Published 9 days ago • 39
