GRNEdit：生成式细化网络中新二元证据视角下的高效通用视频编辑

摘要

基于指令的通用视频编辑旨在通过单一直观的界面统一多样化的编辑操作。现有方法通常依赖资源密集的条件建模，要么使用重量级分支，要么采用昂贵的源拼接。是否存在一种高效的方式来建模编辑意图？为此，我们提出了 GRNEdit，一个轻量级的两阶段框架。GRN 通过位的组合来编码视觉语义，这启发了我们的方法。通过任务特定的微调，我们进一步利用这一表示，将编辑语义重新表述为对单个位的局部保留或翻转决策。因此，源信息被建模为支持观测到的二值状态的逐坐标证据，而 GRN 主干网络仍负责将其全局组合解析为连贯的生成语义。在第一阶段，一个紧凑的编码器将离散的源编码转换为连续的证据信号，GRN 在二值细化过程中吸收这些信号。受无分类器引导中空提示训练的启发，我们进一步赋予空条件一种编辑特定的含义：空指令表示不进行编辑，并通过源重建进行监督。这条恒等路径不仅隐式地加强了对第一阶段的证据利用和内容保持，而且在与编辑状态相同的表示空间中产生一个源保持状态。因此，第二阶段可以直接将每个编辑状态与其源保持对应状态进行比较，并利用它们的差异来修正未解决的目标位决策。仅使用 0.6M 对训练数据，且条件参数占比不到 3%，GRNEdit-2B 和 GRNEdit-8B 在 OpenVE-Bench 上分别取得了 4.03 和 4.18 的分数。其中 2B 模型优于多个 14B 的开源编辑器，而 8B 模型的表现与领先的开源编辑器相当。

English

Instruction-based general video editing seeks to unify diverse editing operations within a single, intuitive interface. Existing approaches often rely on resource-intensive conditioning, using either heavyweight branches or costly source concatenation. Is there any efficient way to model editing intent? Thus, we introduce GRNEdit, a lightweight two-stage framework. GRN inspires our approach by encoding visual semantics through combinations of bits. Through task-specific fine-tuning, we take this representation further and recast editing semantics as local retain-or-flip decisions over individual bits. Source information is consequently modeled as coordinate-wise evidence supporting the observed binary states, while the GRN backbone remains responsible for resolving their global composition into coherent generative semantics. In Stage I, a compact encoder translates discrete source codes into continuous evidence signals, which GRN assimilates throughout binary refinement. Inspired by null-prompt training for classifier-free guidance, we further assign the null condition an editing-specific meaning: an empty instruction denotes no edit and is supervised through source reconstruction. This identity pathway not only implicitly strengthens evidence utilization and content preservation in Stage I, but also produces a source-preserving state in the same representation space as the edited state. Stage II can therefore directly compare each edited state with its source-preserving counterpart and use their discrepancy to revise unresolved target-bit decisions. Trained on only 0.6M pairs with less than 3\% conditioning parameters, GRNEdit-2B and GRNEdit-8B achieve scores of 4.03 and 4.18 on OpenVE-Bench. The 2B model outperforms multiple 14B open-source editors, while the 8B model performs on par with leading open-source editors.