ImageNet 256x256
PulseAugur coverage of ImageNet 256x256 — every cluster mentioning ImageNet 256x256 across labs, papers, and developer communities, ranked by signal.
-
New MEPA architecture enhances visual autoregressive modeling with Mixture of Experts
Researchers have developed MEPA, a novel Mixture of Experts (MoE) architecture designed to improve visual autoregressive modeling. MEPA addresses limitations in multi-scale representation learning by enabling scale-adap…
-
Dual-End Consistency Model accelerates generative AI with improved training and sampling
Researchers have introduced the Dual-End Consistency Model (DE-CM), a novel approach to accelerate generative models like diffusion and flow-based systems. DE-CM addresses training instability and sampling inflexibility…
-
New VRPO method speeds up diffusion transformer training
Researchers have developed a new method called VRPO to improve the training efficiency and image quality of diffusion transformers. This approach replaces static alignment losses with a reinforcement learning objective …