TensorX
返回文献探索

Paper · arXiv 2607.26811

DistillAlign: Coordinating Mode Covering and Mode Seeking in Autoregressive Video Distillation

Jiaxing Li, Kai Zou, Cindy Zhou, Kaichen Huang, Junyao Gao, Zile Wang, Yang Liu, Bin Liu, Bo An, Yangguang Li

93 upvotesJuly 29, 2026arXiv 预印本
AI 摘要

Joint distillation aligns initialization and refinement distributions to improve coverage and diversity in autoregressive video generation.

autoregressive video distillationDistribution Matching DistillationDMDmode-seekingmode-coveringConsistency Distillationreverse-KLprecision and coveragelatent spacejoint distillation

Abstract

Existing autoregressive video distillation methods commonly adopt a Distribution Matching Distillation (DMD)-based multi-stage pipeline. However, they typically decouple the initialization and DMD stages -- which then pursue different target distributions -- and judge the intermediate student mainly by visual scores such as VBench. In this paper, we revisit this design from a distributional perspective. Given the mode-seeking nature of the distribution matching loss, a good initialization should match the mode coverage of the target DMD teacher, rather than merely pursuing high quality. To analyze this, we introduce a distributional evaluation protocol that measures precision and coverage between student and teacher distributions in a shared latent space. It exposes differences hidden by visual scores: some initializations reach high precision but low coverage, leading to suboptimal refinement, while mode-covering ones preserve broader support. Furthermore, even when the target distributions are aligned, DMD's reverse-KL objective can still drive the student toward high-probability teacher regions in late training, reducing coverage and diversity. To address this, we propose joint distillation, which combines DMD's mode-seeking objective with a Consistency Distillation-based mode-covering constraint. Experiments show that our method improves generation quality, coverage, and diversity; notably, even with a Wan-1.3B DMD teacher, it outperforms baselines refined with Wan-14B, underscoring the importance of distributional alignment in autoregressive video distillation.

北京市昌平区探索星信息技术及软件开发工作室

京ICP备2026059466号
DistillAlign: Coordinating Mode Covering and Mode Seeking in Autoregressive Video Distillation | TensorX