AI 中文总结
本文提出近邻正则化合并（PRM），通过在任务向量训练中引入近端惩罚，使顺序LoRA合并更稳健，提升AAA性能，并揭示任务向量的可合并性对持续学习效果的关键作用。
AI 中文摘要
无需重放的无参数高效适配器持续学习可被建模为一系列任务向量写入操作：对于每个新任务，学习一个低秩适配器并将其合并到运行中的模型中。我们提出了近邻正则化合并（PRM），这是对顺序LoRA合并的最小修改，在任务向量训练期间添加近端惩罚，而不改变后续的写入规则。PRM作为一种稳健的任务向量正则化器：在报告的Base->+Prox诊断中，它在多种写入规则、骨干网络和类增量设置下提升了AAA，而其固定系数变体仍能与基于强系数的基线相竞争。机制上，匹配前缀范数控制和近端强度扫描表明，近端训练缩小了任务向量半径，降低了Fisher加权干扰，拓宽了系数平台，并揭示了稳定性-可塑性权衡。综合这些结果，顺序LoRA合并的有效性不仅取决于写入任务向量的多少，还取决于任务向量本身是否被训练为可合并的。
英文摘要
Rehearsal-free continual learning with parameter-efficient adapters can be cast as a sequence of task-vector write-in operations: for each new task, a low-rank adapter is learned and merged into a running model. We propose Proximity Regularized Merging (PRM), a minimal modification to sequential LoRA merging that adds a proximal penalty during task-vector training without changing the subsequent write-in rule. PRM acts as a robust task-vector regularizer: in the reported Base->+Prox diagnostics, it improves AAA across multiple write-in rules, backbones, and class-incremental settings, while its fixed-coefficient variant remains competitive with strong coefficient-based baselines. Mechanistically, matched-prefix norm controls and proximal-strength sweeps show that proximal training shrinks the task-vector radius, lowers Fisher-weighted interference, broadens the coefficient plateau, and exposes a stability-plasticity trade-off. Together, these results suggest that the effectiveness of sequential LoRA merging depends not only on how much of a task vector is written in, but also on whether the task vector itself has been trained to be mergeable.
