ConceptMoE: Adaptive Token-to-Concept Compression for Implicit Compute Allocation Paper • 2601.21420 • Published 5 days ago • 39
Generation Enhances Understanding in Unified Multimodal Models via Multi-Representation Generation Paper • 2601.21406 • Published 5 days ago • 4
Everything in Its Place: Benchmarking Spatial Intelligence of Text-to-Image Models Paper • 2601.20354 • Published 6 days ago • 108
DynamicVLA: A Vision-Language-Action Model for Dynamic Object Manipulation Paper • 2601.22153 • Published 4 days ago • 66