Hacker GitHub

← All repos

chiennv2000/orthrus

View on GitHub ↗ ★ 482 Python

[NeurIPS 2026 Spotlight] Fast, lossless LLM inference via dual-view diffusion decoding.

3 mentions on Hacker News by 2 people: 1 story, 2 comments · first 2026-05-15 · latest 2026-05-16

Mentions per year 3 mentions · busiest year 2026 (3)
20082026
AllStoriesComments NewestTopOldest
  1. And it also looks like the original authors are working on qwen 3.5 too: https://github.com/chiennv2000/orthrus/issues/1#issuecomment...
    comment by dot_treo · 2026-05-16 · in Orthrus-Qwen3: up to 7.8×tokens/forward on Qwen3, identical output distribution
  2. Paper: https://arxiv.org/abs/2605.12825 ; Code+models: https://github.com/chiennv2000/orthrus ; Disclosure: co-author. Idea: Inject a trainable diffusion attention module into each layer of a frozen AR Transformer. Both heads share one KV…
    comment by FranckDernoncou · 2026-05-15 · in Orthrus-Qwen3: up to 7.8×tokens/forward on Qwen3, identical output distribution