NVIDIA Breakthrough: Diffusion Models for Text?
A revolution in AI? The End of Next-Token-Prediction for AI?
How NVIDIA’s SIGMA Generates Continuous Diffusion for text and causal reasoning.
Diffusion for Text is Here: The SIGMA Continuous Diffusion Revolution.
Breaking the LLM Bottleneck: How SIGMA Generates Entire Blocks in Parallel.
All rights w/ authors:
Large Language Continuous Diffusion Models
Zhihan Yang*,†,1,2 Wei Guo*,†,1,3 Jean-Marie Lemercier*,1 Simon Welker*,1 Yonggan Fu1 Mohammad Mahdi Kamani1 Sajad Norouzi1 Julius Berner1, Tomas Geffner1 Karsten Kreis1 Yongxin Chen3 Molei Tao3 John Thickstun2, Pavlo Molchanov1 Ante Juki