DC-Merge：利用方向一致性改进模型合并
DC-Merge: Improving Model Merging with Directional Consistency
对做多任务/多领域视觉模型部署的人，DC-Merge 提供了一种无需额外训练、仅靠奇异值平滑与共享正交子空间投影就能把多个微调模型（含 LoRA）合并成一个的通用后处理手段，可直接用于把多个视觉任务适配器压成单模型。
原文依据 we propose DC-Merge, a method for directional-consistent model merging
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论模型合并旨在将多个任务适配后的模型整合为一个统一模型，并保留各任务的知识。本文指出，这种知识保留的关键在于维持合并后的多任务向量与各单任务向量之间奇异空间的方向一致性。然而，这一一致性常常受到两个问题的破坏：i）任务向量内部能量分布不均衡，少数奇异值主导了总能量，导致合并时语义重要但较弱的成分被忽略；ii）任务向量在参数空间中的几何不一致，使得直接合并会扭曲其潜在的方向几何结构。为应对这些挑战，我们提出 DC-Merge，一种方向一致的模型合并方法。它首先通过平滑奇异值来平衡每个任务向量的能量分布，确保所有知识成分都得到充分表达。随后，这些能量均衡后的向量被投影到一个共享正交子空间上，以最小的重建误差对齐其方向几何。最后，对齐后的向量在共享正交子空间中聚合，并投影回原始参数空间。在视觉与视觉-语言基准上的大量实验表明，DC-Merge 在全量微调与低秩适配两种设置下均持续取得最优性能。实现代码见 https://github.com/Tobeginwith/DC-Merge。
英文原文摘要 1553 字符
Model merging aims to integrate multiple task-adapted models into a unified model that preserves the knowledge of each task. In this paper, we identify that the key to this knowledge retention lies in maintaining the directional consistency of singular spaces between merged multi-task vector and individual task vectors. However, this consistency is frequently compromised by two issues: i) an imbalanced energy distribution within task vectors, where a small fraction of singular values dominate the total energy, leading to the neglect of semantically important but weaker components upon merging, and ii) the geometric inconsistency of task vectors in parameter space, which causes direct merging to distort their underlying directional geometry. To address these challenges, we propose DC-Merge, a method for directional-consistent model merging. It first balances the energy distribution of each task vector by smoothing its singular values, ensuring all knowledge components are adequately represented. These energy-balanced vectors are then projected onto a shared orthogonal subspace to align their directional geometries with minimal reconstruction error. Finally, the aligned vectors are aggregated in the shared orthogonal subspace and projected back to the original parameter space. Extensive experiments on vision and vision-language benchmarks show that DC-Merge consistently achieves state-of-the-art performance in both full fine-tuning and LoRA settings. The implementation code is available at https://github.com/Tobeginwith/DC-Merge.