分离处理LoRA矩阵可避免合并时信息丢失，提升多任务性能。
Crowded in B-Space: Calibrating Shared Directions for LoRA Merging
- 区分LoRA的B和A矩阵，仅校准输出侧的B矩阵。
- 在8个领域基准上平均准确率提升3.4-8.3点，部分超越联合训练。
- 无需数据即可插入现有合并方法，适合多任务模型部署场景。
独立训练的LoRA适配器合并是多任务训练的实用替代方案，但常导致性能下降。现有方法将LoRA更新ΔW = BA视为单一对象，未区分两个矩阵。我们发现，干扰主要源于输出侧矩阵B：不同任务中，B反复使用少量共享方向，而A更具任务特异性。合并后这些共享方向被过度强调，导致任务特异性信息丢失。为此，我们提出Pico（输出空间预合并干扰校准），一种无需数据的方法，在合并前通过缩小过共享方向并重新缩放合并更新来校准B。Pico可直接集成至任务算术、TIES和TSV-M等现有方法。在数学、编程、金融和医疗等八个不同基准上，相比基线方法平均准确率提升3.4-8.3点，并实现最优整体性能。此外，合并后的适配器甚至优于用所有任务数据联合训练的LoRA。结果表明，分别处理两个矩阵能显著提升LoRA合并效果。
原文摘要 · Abstract (English)
Merging separately trained LoRA adapters is a practical alternative to joint multi-task training, but it often hurts performance. Existing methods usually treat the LoRA update $ΔW = BA$ as a single object and do not distinguish the two LoRA matrices. We show that the main source of LoRA merge interference comes from the output-side matrix $B$. Across tasks, $B$ repeatedly uses a small set of shared directions, while $A$ remains much more task-specific. As a result, the merged adapter overemphasizes these shared directions, and task-specific information is lost. We propose Pico (Pre-merge interference calibration in output-space), a data-free method that calibrates $B$ before merge by downscaling over-shared directions and then rescaling the merged update. Pico plugs directly into existing merging methods such as Task Arithmetic, TIES, and TSV-M. Across eight different benchmarks from math, coding, finance, and medical domains, Pico improves average accuracy by 3.4-8.3 points over the corresponding base method and achieves the best overall average performance. Pico also enables merged adapters to outperform the LoRA trained with all task data. These results show that LoRA merging works better when the two LoRA matrices are treated separately.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。