Diffusion models are widely used in many AI applications, but research on efficient inference-time scalability, particularly for reasoning and planning (known as System 2 abilities) has been lacking.

KAIST (President Kwang Hyung Lee) announced on the 20th that a research team led by Professor Sungjin Ahn in the School of Computing has developed a new technology that significantly improves the ...

AMD has officially enabled Stable Diffusion on its latest generation of Ryzen AI processors, bringing local generative AI image creation to systems equipped with XDNA 2 NPUs. The feature arrives ...

AMD's world's first FP16 Stable Diffusion 3.0 Medium model for image generation is now available for AMD Ryzen AI 300 Series ...

Google’s Gemini Diffusion demo didn’t get much airtime at I/O, but its blazing speed—and potential for coding—has AI insiders speculating about a shift in the model wars.

Gero’s Structure-Free AI Model Generates Small Molecules from Protein Sequences. June 25, 2025. ... A Structure-Free Diffusion Language Model for Protein Sequence-Conditioned Ligand Design.” ...

Google previewed a new AI model, Gemini Diffusion, that it claims is “state-of-the-art” on certain coding and math tasks — leveraging parallel generation to achieve low latency. The version ...