Non-autoregressive NC LLM spits tokens 21x faster than text diffusion

  • Posted 4 hours ago by LarsDu88
  • 2 points
https://hazyresearch.stanford.edu/blog/2026-10-12-ncm

0 comments