Termsslug: self-speculative-decoding

Self-Speculative Decoding

自推测解码

Accelerated decoding where the model's own draft head proposes tokens that are then verified in parallel.

Curated Intelligence

1 items · Page 1/1
Marktechpost (@Marktechpost)T2·Models
78 pts
NVIDIA Releases Sigma Diffusion Language Models with 3.46x Faster Decoding

NVIDIA released Nemotron-Labs-Sigma, 14 continuous diffusion language model checkpoints at 3B and 8B that move diffusion out of discrete token space into a learned 16-D embedding space, adding and removing Gaussian noise block by block. Sigma-8B averages 71.16 on NVIDIA's 6-task base suite versus 70.22 for the discrete-diffusion NLD-8B, and self-speculative decoding runs 3.46x faster on average (up to 4.20x) on a single GB200. The trade-off is clear: instruct average of 59.14 trails NLD-8B's 59.92, and AIME25 sits at just 17.00.