Search papers, labs, and topics across Lattice.
Texas A&M University, Carnegie Mellon University
3
2
6
6
Experiments show that MoRE consistently achieves lower perplexity and stronger downstream performance than standard MoEs and state-of-the-art weight-sharing architectures at matched compute and parameter budgets, with only minimal modifications to existing MoE implementations.
A 5-step sampling schedule can deliver nearly the same quality as a 50-step one, slashing inference costs by 90%.
By pausing to "think" with latent diffusion, STAR-LDM achieves superior language understanding, narrative coherence, and controllable generation compared to standard autoregressive models of similar size.
