# Not All Frames Deserve Full Computation: Accelerating Autoregressive Video Generation via Selective Computation and Predictive Extrapolation

Source: [Not All Frames Deserve Full Computation: Accelerating Autoregressive Video Generation via Selective Computation and Predictive Extrapolation](<https://www.sota2.com/research/paper/69d33deb7d0172140ec735bd>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-08-12T05:00:05.228Z

Authors: Hanshuai Cui, Zhiqing Tang, Zhi Yao, Fanshuai Meng, Weijia Jia, Wei Zhao

Published: 2026-04-03T11:34:47.000Z

## Abstract

Autoregressive (AR) video diffusion models enable long-form video generation but remain expensive due to repeated multi-step denoising. Existing training-free acceleration methods rely on binary cache-or-recompute decisions, overlooking intermediate cases where direct reuse is too coarse yet full recomputation is unnecessary. Moreover, asynchronous AR schedules assign different noise levels to co-generated frames, yet existing methods process the entire valid interval uniformly. To address these AR-specific inefficiencies, we present SCOPE, a training-free framework for efficient AR video diffusion. SCOPE introduces a tri-modal scheduler over cache, predict, and recompute, where prediction via noise-level Taylor extrapolation fills the gap between reuse and recomputation with explicit stability controls backed by error propagation analysis. It further introduces selective computation that restricts execution to the active frame interval. On MAGI-1 and SkyReels-V2, SCOPE achieves up to 4.73x speedup while maintaining quality comparable to the original output, outperforming all training-free baselines.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Video Generation on SkyReels DF-1.3B VBench gallery prompts V2 540P (test)](<https://www.sota2.com/research/sota/video-generation-on-skyreels-df-1-3b-vbench-gallery-prompts-v2-540p-test.md>): Video Generation; SkyReels DF-1.3B VBench gallery prompts V2 540P (test); Time (s): 307.34.

- [Video Generation on MAGI-1 4.5B-distill VBench gallery prompts 480x480 (test)](<https://www.sota2.com/research/sota/video-generation-on-magi-1-4-5b-distill-vbench-gallery-prompts-480x480-test.md>): Video Generation; MAGI-1 4.5B-distill VBench gallery prompts 480x480 (test); Time (s): 310.16.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2604.02979>)

[PDF](<https://arxiv.org/pdf/2604.02979.pdf>)
