{"slug": "joint-alignment-and-distillation-for-video-generation-via-sample-guided-matching", "title": "Joint Alignment and Distillation for Video Generation via Sample-Guided Distribution Matching", "summary": "Researchers propose DM-Align, a unified single-stage optimization framework for video generation that combines distillation and preference alignment via sample-guided distribution matching, eliminating the need for multi-step reward evaluation and complex ODE-SDE conversions. Experiments across multiple foundational video models show it outperforms standalone variants and sequential two-stage pipelines in both distillation quality and preference alignment.", "body_md": "arXiv:2609.04283v1 Announce Type: new \nAbstract: Aligning video generative models to human preferences heavily relies on Reinforcement Learning (RL), which suffers from extensive computational overhead. Existing workflows typically treat RL and distillation as disconnected stages: applying RL before distillation incurs prohibitive computational costs, whereas applying RL after distillation frequently leads to model collapse. To overcome these limitations, we propose a unified, single-stage optimization framework grounded in Distribution Matching (DM). In the standard DM framework, distillation updates the model via a gradient direction that minimizes the gap between the real and fake models, guiding generations toward clarity and high fidelity. Building upon this, we introduce DM-Align, which derives a complementary gradient direction to guide the model toward human-preferred samples. Inspired by DPO and GRPO, our method leverages the distributional gap -- formulated from either preference pairs or intra-group exploration -- to directly construct this preference-guided gradient. By synergizing these two gradient directions, our approach eliminates the need for multi-step reward evaluation and complex ODE-SDE conversions inherent in traditional RL. Comprehensive experiments across multiple foundational video models demonstrate that this sample-guided framework robustly enhances both distillation quality and preference alignment, consistently outperforming both standalone variants and sequential two-stage pipelines.", "url": "https://wpnews.pro/news/joint-alignment-and-distillation-for-video-generation-via-sample-guided-matching", "canonical_source": "https://arxiv.org/abs/2609.04283", "published_at": "2026-09-07 04:00:00+00:00", "updated_at": "2026-09-07 04:26:08.657853+00:00", "lang": "en", "topics": ["artificial-intelligence", "generative-ai", "machine-learning"], "entities": ["DM-Align", "arXiv"], "alternates": {"html": "https://wpnews.pro/news/joint-alignment-and-distillation-for-video-generation-via-sample-guided-matching", "markdown": "https://wpnews.pro/news/joint-alignment-and-distillation-for-video-generation-via-sample-guided-matching.md", "text": "https://wpnews.pro/news/joint-alignment-and-distillation-for-video-generation-via-sample-guided-matching.txt", "jsonld": "https://wpnews.pro/news/joint-alignment-and-distillation-for-video-generation-via-sample-guided-matching.jsonld"}}