Recent video diffusion models (VDMs) synthesize visually convincing clips, yet still drop entities, mis-bind attributes, and weaken the interactions specified in the prompt. Representation-alignment objectives such as VideoREPA and MoAlign improve fine-grained text following by distilling spatio-temporal token relations from a frozen visual foundation model, but their pairwise supervision budget is allocated by visual or motion cues rather than by how relevant each pair is to the prompt. We present SARA, Semantically Adaptive Relational Alignment, which keeps token-relation distillation (TRD) on a frozen VFM target and adds a text-conditioned saliency that decides which token pairs carry supervision. A lightweight Stage~1 aligner is trained with per-entity SAM~3.1 mask supervision and an InfoNCE regulariser, and its continuous saliency is fused into TRD through a pair-routing operator that assigns each token pair a weight whenever either of its two endpoints is salient, thereby routing supervision toward subject-subject and subject-background pairs and away from background-background ones. In the Wan2.2 continual-training setting, SARA improves both text alignment and motion quality over SFT, VideoREPA, and MoAlign on a 13-dimension VLM rubric, on the public VBench benchmarks, and in a blind user study. Project page: https://saradit.github.io/.

翻译：摘要：近期视频扩散模型能够合成视觉上令人信服的片段，但仍会出现实体遗漏、属性误绑定及提示词中指定交互作用弱化等问题。VideoREPA与MoAlign等表征对齐目标通过从冻结视觉基础模型中提取时空令牌关系来改进细粒度文本跟随性能，但其成对监督预算由视觉或运动线索分配，而非依据各令牌对与提示词的相关性。本文提出语义自适应关系对齐方法SARA，该方法保持对冻结VFM目标的令牌关系蒸馏，并加入基于文本的显著性机制来决定需监督的令牌对。轻量级Stage-1对齐器通过逐实体SAM 3.1掩码监督与InfoNCE正则化器训练，其连续显著性通过成对路由算子融入TRD：当令牌对中任一端点显著时，该算子为其分配权重，从而将监督导向主体-主体与主体-背景对，而减少背景-背景对的监督。在Wan2.2持续训练设置中，SARA在13维VLM评分体系、公开VBench基准测试及盲用户研究中，于文本对齐与运动质量两方面均超越SFT、VideoREPA及MoAlign。项目页面：https://saradit.github.io/。