2DSD Speculative Decoding Transfers to Diffusion and Flows, Up to 8x Throughput Gain
LucaAmb · x · 2026-11-20
Arnav Chavan notes 2DSD opens a new axis for speculative decoding: results shown on Looped AR transfer to any looped architecture, including Looped AR, Diffusion, and Flows, with up to 8x throughput gains.
Checkpoints are live in a Hugging Face collection and code is open-sourced on GitHub, making looped AR models considerably more practical.
Related event: 2DSD Speculative Decoding Speeds Up Looped LMs Up to 8x(2 posts)→
More from Infra
- Starlink Mini enables remote water gate control in mountainous Japan; all 6 tests successful — XFreeze · 2026-11-20
- Tencent Youtu's LightningMoBA speeds long-context attention 1.2-1.6x without quality loss — tencent · 2026-11-20
- Jensen Huang's personal AI clout won't transfer to Nvidia's next CEO, analyst says — FinanceYF5 · 2026-11-20
- Europe's AI test happens inside the factory: sovereign compute must fit installed operations — ingliguori · 2026-11-20
- Mubit + Neon: agent middleware cuts frontier-model tokens by 73% — JeremyCMorgan · 2026-11-20
- 2DSD Speculative Decoding Speeds Up Looped LMs up to 7.64x With Lower KV Memory — LucaAmb · 2026-11-20