Speculative Forcing for Efficient Autoregressive Video Generation
Few-step autoregressive video generators enable streaming generation, yet they run every token through the final denoising step, regardless of how much refinement it needs. We find this uniform computation is often unnecessary: many tokens already closely match their eventual full-step outputs in the initial steps, while others require refinement until the end. We introduce Speculative Forcing, a framework that treats intermediate clean estimates as drafts and predicts when individual tokens can be safely accepted early. Once accepted, a token is frozen as its final output and undergoes no further denoising, while its cached keys and values remain available as context for the remaining to...
| Authors | Hyunwook Choi, Seungryong Kim, Dongyoon Han et al. |
|---|---|
| Primary category | cs.CV |
| Published | 15 October 2026 |
| Last updated at source | 15 October 2026 |
Recent changes
- 16 Oct 2026 Description: Few-step autoregressive video generators enable streaming generation, yet they run every token through the final denoising step, regardless of how much refinement it needs. We find this uniform computation is often unnecessary: many tokens → Few-step autoregressive video generators enable streaming generation, yet they run every token through the final denoising step, regardless of how much refinement it needs. We find this uniform computation is often unnecessary: many tokens
- 16 Oct 2026 Topics added cs.CV removed hf-daily-papers, autoregressive video generators, denoising, speculative forcing, clean estimates, verifier, frozen generator, intermediate features
Changes detected by Silytics between refreshes of the source.
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 16 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.